This is the peer reviewed version of the following article: Robertson C et al. Research Synthesis
Methods2014;5(3):200-211,
which
has
been
published
in
final
form
at
DOI:
10.1002/jrsm.1102. This article may be used for non-commercial purposes in accordance With
Wiley Terms and Conditions for self-archiving
Title: Practicalities of using a modified version of the Cochrane Collaboration risk of bias
tool for randomised and non-randomised study designs applied in a health technology
assessment setting
Introduction
Systematic literature reviews are recognised as providing important information to health
care practitioners and policy makers by summarising the best available research evidence on
given health care interventions. However, inclusion of methodologically flawed studies can
exaggerate treatment effects, which in turn leads to biased conclusions. Assessing risk of
bias of individual included studies is therefore considered essential to ensure the validity of
review conclusions (Juni et al., 2001; Wood et al., 2008).
The Cochrane Collaboration Bias Methods group developed a specific tool between 2005 and
2007 for assessing risk of bias. The tool was designed to allow a domain-based evaluation of
the internal validity of randomised controlled trials (RCT). Raters must judge whether a trial
is at high, low or unclear risk of bias across six domains: sequence generation and allocation
concealment (selection bias), blinding of participants, healthcare providers and outcome
assessors (performance and detection bias), incomplete outcome data (attrition bias), selective
outcome reporting and ‘other’ potential sources of bias that are specific to the review.
Assessments are made on the basis of the trial report and any other supporting evidence, such
as the trial protocol (Higgins et al., 2011).
In some cases it will not be possible to answer the review question with RCT evidence alone,
for example, where RCT evidence is lacking, has limited information on important outcomes
or where it would be unethical to carry out randomisation (Britton et al., 1998; Deeks et al.,
2003). Non-randomised studies (NRS) may provide valuable evidence in the absence of RCT
data and may often be the only source of evidence about the effects of interventions on longterm outcomes, rare events and adverse effects (Deeks et al., 2003). The NRS design is more
susceptible to the influence of certain biases, however; selection bias and confounding in
1
particular. Furthermore, judging bias at study level and outcome domains in NRS can be
particularly challenging. For example, a study judged as being at low risk of bias for
confounding for a given outcome could be so highly biased at a study level domain that it
becomes questionable whether the study can safely be judged as being at low risk of bias for
confounding. Consequently, conducting risk of bias assessments in reviews of NRS is likely
to be more problematic and complex than for reviews including RCT evidence only. The
Cochrane Non-Randomised Studies Methods Group (NRSMG) issued draft guidance on
including NRS in systematic reviews and provided a modified Cochrane Collaboration risk of
bias tool to include outcome specific confounders, the nature of which will vary depending
on individual review topics (Reeves et al., 2010, Barney Reeves, personal communication,
2010). The modified tool could be used for both RCTs and NRS. To the best of our
knowledge, there has been no published empirical assessment of the practicalities of using a
NRS risk of bias tool.
Therefore, we aimed to evaluate the practicalities of using the tool in a systematic review of
laparoscopic versus robotic prostatectomy that included RCT and non-randomised
comparative cohort studies. Our objectives were:
 To assess time to complete risk of bias assessment; to judge if its use was feasible within
the time constraints of a health technology assessment project
 To assess inter-rater agreement for risk of bias assessment using the modified Cochrane
Risk of bias tool; to assess reliability of the tool
 To explore the association between risk of bias and treatment effect size; to see if
restriction to low risk of bias studies resulted in greater confidence in effect estimates.
Methods
Example dataset
The review described (Ramsay et al., 2012) was tasked with determining whether complete
removal of the prostate (radical prostatectomy) was best achieved using standard
laparoscopic (keyhole) surgery or robot-assisted laparoscopic surgery, known as robotic
prostatectomy). Our sample comprised full text primary reports included in a systematic
2
review comparing laparoscopic versus robotic prostatectomy for localised prostate cancer
(n=48). Of these, one report was an RCT and the remainder were comparative cohort studies.
Tool Development
The risk of bias tool was adapted from that presented at a workshop run by the NRSMG
at the 15th Annual Meeting of UK and Ireland-based Contributors to The Cochrane
Collaboration. We included the standard risk of bias domains and identified potential
topic-specific, outcome focused confounders. Our risk of bias assessment focused on the
following review specific outcomes, which were developed for the a priori protocol for
the main review project:
 peri-operative safety - complications and adverse events arising from robotic or
laparoscopic prostatectomy surgery including blood transfusion, anastomotic leak, bladder
neck contracture, wound infection, organ injury, ileus, deep vein thrombosis and
pulmonary embolism
 urinary dysfunction - duration and degree of incontinence following radical prostatectomy
 erectile dysfunction – duration and degree of erectile dysfunction following radical
prostatectomy
 efficacy - rate of positive margin in resected specimen (predominantly); biochemical
(PSA) recurrence; need for further cancer treatment; disease free survival, defined as
absence of clinically detectable disease; survival; mortality.
For each of these outcomes we identified relevant potential confounders by reviewing the
existing literature in the prostate cancer field. Clinical members of the Project Working
Group then compiled a list of relevant confounders for each outcome. The Project Advisory
Group reviewed the generated list of confounders for completeness.
Confounders were
ranked in order of their prognostic importance by our clinical experts to assist reviewers’
decision making processes. These were arranged in a hierarchy in the tool with the most
influential confounder appearing first.
Differences of opinion were resolved through
discussion amongst clinical expert members of the Project Working and Advisory Groups
and through arbitration by the clinical chief investigator (RP).
The topic-specific
confounders related to specific outcomes are shown in our modified tool (Appendix 1).
3
Lastly, we added domains to assess whether study authors had prepared an a priori protocol
and analysis plan. (Higgins et al 2011)
Five reviewers independently assessed the risk of bias of included English full text studies.
Reviewers were than randomly assigned in pairs to give a total of six review pairings.
Random assignment was achieved via computer generated random numbers. Any differences
in assessment or issues of uncertainty were resolved by discussion and consensus between the
reviewers. The review team had 2-13 years’ collective experience of conducting systematic
reviews. CRa is a Professor of Health Services Research and Health Care Assessment
Programme Director and has extensive experience of designing and conducting systematic
reviews and clinical trials. GM is a Senior Research Fellow, and lead of a large Evidence
Synthesis team. TG, PS and CRo are all Research Fellows and lead reviewers of previous
systematic reviews. None of the review team are clinicians but received training in natural
history of and treatments for prostate cancer in the form of workshop presentations by the
clinical members of the project team prior to commencing the review. Reviewers also
received clinical guidance throughout the risk of bias assessment process.
The risk of bias assessment was summarised at the study level using judgements
incorporating individual outcomes as well as study level risk of bias domains. Individual
outcomes were categorised as high risk of bias, low risk of bias or unclear risk of bias. For
example, in evaluating confounding for peri-operative safety for one included study (Hu et
al., 2006) reviewer 1 judged this domain at high risk, reviewer 2 judged it as low risk and
after discussion both returned the agreement of low risk of bias. Where confounders were
balanced between groups, or were unbalanced but authors reported using an appropriate
statistical method to control for confounding, a judgement of low risk of bias was made.
Where confounders were not balanced and no attempt was made to control for the imbalance,
the study was judged to be at high risk of bias for this domain. We decided to use the
response options low, high and unclear risk of bias, rather than adopting a 5-point scale
approach, following guidance from another review group. These reviewers indicated that
they found the numerical scale to be ambiguous and experienced difficulty reaching scoring
agreement. (Academic Urology Unit, University of Aberdeen, personal communication
October 2010)
Data analysis
4
We calculated Kappa statistics to assess inter-rater agreement between reviewers for each risk
of bias domain for each study with 0–0.2 as slight agreement, 0.21–0.4 as fair, 0.41–0.6 as
moderate, 0.61–0.8 as substantial, and 0.81–1 as perfect (Landis and Koch, 1977). The
categories were weighted to reflect higher disagreement between the two clear categories of
low and high risk compared to lower weighting for disagreement between either high or low
and unclear judgements. For example, if reviewer 1 scored high and reviewer 2 scored low
this would be more heavily weighted than if reviewer 2 had scored unclear.
Any
disagreements were resolved by consensus or arbitration by a third party. Kappa statistics
were calculated after reviewers had reached consensus. A scoring scale approach based on
design features was avoided as this has been reported to be inaccurate concerning the
direction of bias and can include items that are unrelated to the internal validity of a study
(Juni et al., 1999).
To explore the relationship between treatment effect size and risk of bias, where there were at
least two studies allowing robotic and laparoscopic comparison for a given outcome, the
treatment effect size was explored by repeating data analyses of all studies including only
data from studies assessed as low risk of bias. The times taken to reach individual and joint
decisions were recorded.
Results
Overall assessment of risk of bias
Of the 48 studies 24 (50%) were judged to be at high overall risk of bias. Eleven studies
(23%) were judged as unclear. Only the RCT (Guazzoni et al., 2006) was judged to be at low
risk of bias for sequence generation and one NRS (Touijer et al., 2007) was judged to be at
low risk for allocation concealment. All other studies were high risk or unclear for these two
key domains.
Risk of bias for reported outcomes
The risk of bias assessment for our chosen main outcomes of efficacy (predominantly
surgical margins status), peri-operative adverse events, urinary incontinence, and erectile
dysfunction are summarised in Figures 1 to 5.
5
Peri-operative safety: Thirty-five studies were assessed for risk of bias for reporting of perioperative adverse events. Of these studies, 11 (31%) were judged to be at low risk of bias for
confounding factors.
Urinary dysfunction: Twenty-three studies were assessed for risk of bias for reporting of
urinary incontinence outcomes and ten (43%) were considered to be at low risk of bias for
confounding factors.
Erectile Dysfunction: Twenty studies were assessed for risk of bias for reporting of erectile
dysfunction. Nine (39%) were considered to be at low risk of bias for confounding.
Efficacy: Thirty-seven reports were assessed for risk of bias for efficacy outcomes. The
majority of these, 30 (81%), were considered to be at low risk of bias for confounding
factors.
Risk of bias assessment
The median (range) time for individual assessment was 30 minutes (10 to 49 minutes). The
median (range) time for reaching agreement between reviewers was 10 minutes (2 to 38
minutes). Therefore an average time of 40 minutes was required to assess risk of bias in each
study. Table 1 shows the analysis of inter-rater agreement for the individual domains of the
risk of bias tool. Inter-rater agreement was slight for the majority of domains. Perfect
agreement was reached for only one outcome specific domain, blinding of peri-operative
safety, and moderate agreement was reached in assessing confounding for efficacy outcomes
only. In assessing overall risk of bias, inter-rater agreement was fair with standard and
weighted Kappa values of 0.34 and 0.35 respectively.
We examined the qualitative data that reviewers recorded while making their assessments in
the description column of the tool, to provide insight into the nature of the disagreements
between reviewers. We grouped data under four themes: problems using the tool, review
topic disagreement, disagreements due to NRS design and other. Table 2 shows the number
of disagreements encountered from the total number of judgements that were required for all
included studies for each theme. Tool-specific problems were encountered when reviewers
differed in their interpretation of risk due to non-reporting issues within the included studies.
This caused disagreement in just over half of all judgements (52%). To illustrate this, the
6
study conducted by Artibani et al (2003) failed to report details for allocation concealment, or
whether an a priori protocol or analysis plan were put in place. One reviewer recorded high
risk assessment for these domains whereas the second reviewer assessed them as unclear.
Review topic disagreements were due to difference in judging whether individual
confounders were balanced, or in judging the prognostic influence of individual confounders
for determining risk of confounding at the outcome level for each of our considered
outcomes. For example, for the study conducted by Bhayani et al (2003) both reviewers
assessed surgeon experience as low risk of bias and co-morbidity and prostate size as unclear
when assessing risk of bias for peri-operative safety. One reviewer recorded a judgement of
low risk of bias for confounding for this outcome, whereas the other reviewer assessed the
outcome as unclear. As the majority of included studies had a NRS design, this caused
differences in subjective interpretation for other and overall risk of bias domains in 48% of
all judgements. One reviewer, for example, assessed all non-randomised studies as being at
high overall risk of bias, while other reviewers tended to be less conservative. ‘Other’
disagreements were due to differences in interpreting semantics of reported text (21%) or
misreading of reported text (6%). It was also evident that, reviewers used the unclear
response option when it was difficult to decide between low and high risk of bias, rather than
due to inadequate information or unclear wording in the study text.
The effect estimate for positive margin for all studies was 0.61 (95% credible interval (CrI)
0.46 to 0.83). The effect estimate for positive margin for studies assessed as at overall low
risk of bias was 0.73 (95% CrI 0.29 to 1.75). Figure 6 shows the treatment effect size and CrI
for all studies and for studies judged as low risk of bias only.
Discussion
We identified few studies that were at low risk of bias for confounding for all outcomes and
few studies that were at overall low risk of bias. This is unsurprising as NRS designs are
inherently more prone to bias than the gold standard RCT design and include differing
attributes which can have varying influences on the potential for bias. That we found only
fair inter-rater agreement between the reviewers illustrates that risk of bias can be interpreted
in different ways by different people. This is particularly likely in the newly developing
methodological area of summarising NRS study design features where terms may be
unfamiliar and impact of study features on magnitude of biases in non-randomised studies is
7
unknown. For example, the term ‘prospective study’ and the term ‘retrospective study’ are
particularly ambiguous.
‘Prospective study’ should imply that all design aspects were
planned, including hypothesis generation, recruitment of participants, baseline data collection
and outcome data collection. In practice, how prospective a study is can often be unclear,
since some aspects of a study can be prospective, such as hypothesis generation and
determination of outcomes, while others are retrospective, such as length of stay data
collection from hospital records (Reeves et al 2011). It is also less likely that study protocols
will be available to support assessments of bias with NRS than is the case for RCTs (Reeves
et al 2011). In this case the reviewer must depend on the study report, which can further
introduce an element of subjectivity when making risk of bias evaluations as the level of
reporting for NRS is often poor. Furthermore, reviewers with little experience of assessing
NRS could be inclined to view all such study designs as being at high risk of bias purely
because they are not of the gold standard RCT design.
In making our judgements for risk of bias due to confounding for each outcome we found that
few studies reported making any statistical adjustments for confounding, therefore
judgements were made based on whether confounding factors were balanced between cohort
groups. Although all reviewers received training and clinical guidance, none are clinicians.
It is possible that we would have achieved better agreement for this domain had we elicited
more detailed a priori decision rules regarding prognostic weighting of confounders. This
demonstrates the importance of having an a priori hierarchy of confounders to aid decision
making and highlights the need for a detailed knowledge of epidemiology and/or statistical
knowledge of appropriate methods for controlling for confounding when planning risk of bias
assessments of NRS. We note, however, that in every case it was possible for the two
reviewers to agree a judgement, suggesting that whilst there was marked disagreement by
single assessment, double assessment appeared to be more robust.
The additional time required to identify key confounders and outline decision guides for
evaluating bias for this domain will vary by review topic but should be factored into the tool
development stage of any systematic review. The process of conducting the risk of bias
assessment and additional analyses was time consuming, with time to reach overall
agreement almost doubling times reported for assessments of RCTs (Hartling et al., 2009;
Hartling et al., 2011). These additional time constraints may limit feasibility for using the
tool within projects with short timelines.
8
In making our judgements of high, low and unclear risk of bias, both for individual domains
and overall study level risk, reviewers often discussed using different subjective criteria to
make a judgement of high or unclear risk of bias. Including a fourth level of ‘moderate’ risk
of bias would have been useful in these cases and could have improved our inter-rater
reliability. This suggestion is in keeping with a proposal to introduce an additional response
category, raised by the NRSMG at the 15th Annual Meeting of UK and Ireland Contributors,
2010. Including a ‘not reported’ response option could also improve disagreements arising
from non-reporting issues.
Restricting our analyses to low risk of bias studies demonstrated more conservative effect
estimates in our systematic review, although reduction in bias incurred greater imprecision.
Conducting a sensitivity analysis such as this was very helpful in discussing the effect of bias
on meta-analysis results in our systematic review; however, we do not recommend replacing
primary meta-analysis results based on these risk of bias assessments due to this loss of
precision and the poor agreement achieved between reviewers in making assessments. If a
robust risk of bias assessment for NRS could be undertaken, then meta-analysis restricted to
low risk of bias studies may be desirable.
Our evaluation has demonstrated that subjectivity is an important factor in carrying out risk
of bias assessments. A limitation of our evaluation is therefore the number of reviewers
involved. All were experienced in conducting systematic reviews but varied in years of
experience. This also highlights the need for clear, detailed guidance for reviewers of
varying levels of experience to further optimise the reliability of the tool. The sample
included prostatectomy studies only so results may not be generalisable to reviews including
non-randomised studies in other topic areas. Given the practical difficulties, we would not
advocate that the tool in this paper should be adopted. There is, however, a clear need for a
risk of bias tool for non-randomised studies, although more detailed guidance for reviewers is
urgently required. An additional level of risk and further validation could improve the
reliability of the current tool.
Acknowledgement and disclaimer
9
This project was funded by the NIHR Health Technology Assessment Programme: Health
Technology Assessment 2012; Vol. 16: No. 41. See the HTA programme website for further
project information.
This report presents independent research commissioned by the National Institute for Health
Research (NIHR). The views and opinions expressed therein are those of the authors and do
not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA Programme or the
Department of Health.
Copyright
“© Queen’s Printer and Controller of HMSO 2012. This work was produced by the UK
Robotic and Laparoscopic Prostatectomy HTA Study Group under the terms of a
commissioning contract issued by the Secretary of State for Health. This journal issue may
be freely reproduced for the purposes of private research and study and extracts (or indeed,
the full report) may be included in professional journals provided that suitable
acknowledgement is made and the reproduction is not associated with any form of
advertising. Applications for commercial reproduction should be addressed to NETSCC,
HTA.”
References
Artibani W, Grosso G, Novara G, Pecoraro G, Sidoti O, Sarti A, et al. 2003. Is laparoscopic
radical prostatectomy better than traditional retropubic radical prostatectomy? An analysis of
peri-operative morbidity in two contemporary series in Italy. Eur Urol 44:401–6.
Bhayani SB, Pavlovich CP, Hsu TS, Sullivan W, Su LM. 2003. Prospective comparison of
short-term convalescence: laparoscopic radical prostatectomy versus open radical retropubic
prostatectomy. Urology 61: 612–16.
Britton A, McKee M, Black N, McPherson K, Sanderson C, Bain C. 1998. Choosing
between randomised and non-randomised studies: a systematic review. Health Technol
Assessment 2:13.
10
Deeks JJ, Dinnes J, D’Amico R, Sowden AJ, Sakarovitch C, Song F, et al. 2003. Evaluating
non-randomised intervention studies. Health Technol Assess 7: 27.
Hartling L, Bond K, Vandermeer B, Seida J, Dryden DM, et al. 2011. Applying the Risk of
Bias Tool in a Systematic Review of Combination Long-Acting Beta-Agonists and Inhaled
Corticosteroids
for
Persistent
Asthma.
PLoS
ONE
6
(2):
e17242.
doi:10.1371/journal.pone.0017242
Hartling L, Ospina M, Liang Y, Dryden DM, Hooton N, Krebs Seida J, Klassen TP. 2009.
Risk of bias versus quality assessment of randomised controlled trials: cross sectional study.
BMJ 339: b4012
Higgins JPT, Altman DG, Sterne JAC (editors). Chapter 8: Assessing risk of bias in included
studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of
Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011.
Available from www.cochrane-handbook.org
Hu JC, Nelson RA, Wilson TG, Kawachi MH, Ramin SA, Lau C, et al. 2006 Perioperative
complications of laparoscopic and robotic assisted laparoscopic radical prostatectomy. J Urol
175: 541–6.
Jüni P, Altman DG, Egger M. 2001. Systematic reviews in health care: Assessing the quality
of controlled clinical trials. BMJ 323: 42-46.
Juni P, Witschi A, Bloch R, Egger M. 1999. The hazards of scoring the quality of clinical
trials for meta-analysis. JAMA 282: 1054–60
Landis JR, Koch GG. 1977. The measurement of observer agreement for categorical data.
Biometrics 33: 159–747.
Ramsay C, Pickard R, Robertson C, Close A, Vale L, Armstrong N, et al. 2012. Systematic
review and economic modelling of the relative clinical benefit and cost-effectiveness of
laparoscopic surgery and robotic surgery for removal of the prostate in men with localised
prostate cancer. Health Technol Assess 16: 41.
11
Reeves BC, Wells G, Shea B. 2010. Classifying non-randomised studies (NRS) and the
assessing the risk of bias for a systematic review. Workshop at the 15th Annual Meeting of
UK and Ireland Contributors to The Cochrane Collaboration, Cardiff.
Reeves BC, Deeks JJ, Higgins JPT and Wells GA on behalf of the Cochrane NonRandomised Studies Methods Group. Including non-randomised studies. Chapter 13
[document on the internet]. In: Cochrane Handbook for Systematic Reviews of Interventions
Version 5.1.0 [updated March 2011].The Cochrane Collaboration; 2011 [accessed March
2011]. http://www.cochrane-handbook.org
Wood L, Egger M, Gluud LL, Schulz K, Jüni P, Altman DG, Gluud C, Martin RM, Wood
AJG, Sterne JAC. 2008. Empirical evidence of bias in treatment effect estimates in controlled
trials with different interventions and outcomes: meta-epidemiological study. BMJ 336: 601605.
12
Table 1 inter-rater agreement for individual risk of bias domains
Domain
Sequence generation
Allocation
concealment
Confounding
Blinding
Incomplete outcome
data
Selective reporting
Standard
Kappa
0.50
0.26
Standard 95% CI
Error
0.07
0.36, 0.64
0.12
0.03, 0.50
Weighted
Kappa
0.66
0.33
Standard 95% CI
Error
0.10
0.46, 0.87
0.11
0.11, 0.55
Peri-operative safety
0.09
0.12
0.03
0.13
Urinary dysfunction
0.11
0.15
0.08
0.17
Erectile dysfunction
0.22
0.16
0.16
0.19
Efficacy
Peri-operative safety
Urinary dysfunction
0.43
1.00
0.13
0.12
0.15
0.51
1.00
0.09
0.14
0.16
Erectile dysfunction
0.09
0.16
0.06
0.17
Efficacy
Peri-operative safety
0.30
0.38
0.12
0.13
0.30
0.40
0.12
0.13
Urinary dysfunction
0.02
0.15
0.01
0.16
Erectile dysfunction
-0.09
0.06
0.01
0.13
Efficacy
-0.05
0.11
-0.06
0.12
Peri-operative safety
Urinary dysfunction
0.00
-0.10
0.00
0.15
0.00
-0.11
0.00
0.17
-0.15,
0.33
-0.18,
0.40
-0.09,
0.53
0.20, 0.67
-0.16,
0.42
-0.22,
0.40
0.07, 0.54
0.13, 0.64
-0.27,
0.31
-0.21,
0.03
-0.27,
0.17
0.00
-0.39,
0.19
-0.23,
0.29
-0.25,
0.14
-0.21,
0.53
0.24, 0.78
-0.22,
0.40
-0.27,
0.39
0.07, 0.54
0.15, 0.66
-0.30,
0.32
-0.24,
0.27
-0.30,
0.18
0.00
-0.44,
0.22
13
Erectile dysfunction
0.12
0.17
Efficacy
0.22
0.13
Other bias
0.16
0.11
A priori protocol
0.18
0.11
A priori analysis
0.01
0.11
Overall risk of bias
0.34
0.10
-0.21,
0.45
-0.04,
0.48
-0.06,
0.38
-0.04,
0.40
-0.21,
0.23
0.14, 0.54
0.05
0.18
0.23
0.13
0.07
0.11
0.21
0.11
0.10
0.10
0.35
0.11
-0.30,
0.40
-0.03,
0.49
-0.15,
0.29
-0.01,
0.43
-0.10,
0.30
0.13, 0.25
14
Table 2
Number of disagreements encountered by number of judgements required
during risk of bias assessment
Coding Theme
n/N (%)
Different interpretation of bias due to non-reporting
25/48 (52%)
Different interpretation of semantics/meaning of reported text
20/48 (42%)
Different interpretation of other/overall bias due to NRS design
10/48 (21%)
One reviewer missed information during 1st reading
3/48 (6%)
Different interpretation of Individual confounders
46/720 (6%)
Peri-operative safety
14/144
Urinary dysfunction
14/192
Erectile dysfunction
8/192
Efficacy
10/192
Different interpretation of prognostic impact of individual confounders for
8/192 (3%)
outcome level decisions
Peri-operative safety
4/48
Urinary dysfunction
2/48
Erectile dysfunction
0/48
Efficacy
2/48
15
Appendix 1
Risk of bias tool
Laparoscopic versus robotic prostatectomy for localised prostate cancer
Assessor initial:
Date evaluated:
Study ID:
Judgement Description (quote from
a
paper, or describe key
information)
Item
1.
Sequence
generation
2.
Allocation
concealment
3a. Confounding
b
Outcome 1
(peri-op safety)
Confounders
balanced b
Surgeon experience
Co-morbidity (ASA/Charlson
score)
Prostate size
3b. Confounding Outcome 2
b
(urinary dysfunction)
Confounders
balanced b
Surgeon experience
Age
Neurovascular bundle
excision
Anastomotic stricture
3c.
Confounding b
Outcome 3
(erectile dysfunction)
Confounders
balanced b
Pre-op dysfunction/status
Neurovascular
excision
bundle
16
Judgement Description (quote from
a
paper, or describe key
information)
Item
Surgeon experience
Age/Co-morbidity
3d.
Confounding b
Outcome 4
(efficacy)
Confounders
balanced b
Gleason score balanced at
baseline
Surgeon experience
PSA score balanced at
baseline
4a. Blinding?
4b. Blinding?
Clinical b1 tumour
stage/nodal stage balanced
at baseline
Outcome 1
(peri-op safety)
Outcome 2
(urinary dysfunction)
4c. Blinding?
Outcome 3
(erectile dysfunction)
4d. Blinding?
Outcome 4
(efficacy)
5a. Incompl.
outcome
data
addressed?
5b. Incompl.
outcome
data
addressed?
5c. Incompl.
outcome
data
addressed?
5d. Incompl.
outcome
Outcome 1
(peri-op safety)
Outcome 2
(urinary dysfunction)
Outcome 3
(erectile dysfunction)
Outcome 4
(efficacy)
17
Judgement Description (quote from
a
paper, or describe key
information)
Item
6a.
6b.
6c.
6d.
data
addressed?
Free of
selective
reporting?
Free of
selective
reporting?
Free of
selective
reporting?
Free of
selective
reporting?
Outcome 1
(peri-op safety)
Outcome 2
(urinary dysfunction)
Outcome 3
(erectile dysfunction)
Outcome 4
(efficacy)
7.
Free of other bias?
8.
A priori protocol? c
9.
A priori analysis plan? d
a
b
Some items on low/high risk/unclear scale, some on yes/no/unclear scale.
For all items, record “unclear” if inadequate reporting prevents a judgement
being made.
Confounders listed by order of importance (high to low importance) Based on list
of confounders considered important at the outset and defined in the protocol for
the review (and assessment against worksheet - optional)
Low risk:
4 balanced = low risk
3 balanced, 1 unbalanced = low risk
3 balanced, 1 unclear = low risk
2 balanced, 1 unbalanced, 1 unclear = low risk
2 balanced, 2 unclear = low risk
18
High risk:
4 unbalanced = high risk
3 unbalanced, 1 balanced = high risk
3 unbalanced, 1 unclear = high risk
2 unbalanced, 2 balanced = high risk
2 unbalanced, 1 balanced, 1 unclear = high risk
2 unbalanced, 2 unclear = high risk
Unclear:
4 unclear = unclear
3 unclear, 1 balanced = unclear
3 unclear, 1 unbalanced = unclear
b1
or pathological stage balanced in absence of clinical stage information.
NB. If confounders are imbalanced but adjusted for in the analysis, the
imbalance is no longer a serious concern for risk of bias.
c
Did the researchers write protocol defining the study population, intervention and
comparator, primary and other outcomes, data collection methods, etc. in
advance of starting the study?
d
Did the researchers have an analysis plan defining the primary and other
outcomes, statistical methods, subgroup analyses, etc. in advance of starting the
study?
General Decision Rules
Where a paper does not report details of confounders/other source of bias this
should be judged as unclear.
Where a paper does not report a considered outcome this should be judged as not
applicable.
Allocation concealment should be judged as high risk of bias if groups are allocated
by factors such as surgeon decision, patient preference.
Allocation by
hospital/institution = low risk. Where no details are given, judge as unclear.
19
Surgeon experience: Assume surgeons performing open prostatectomy are
experienced unless stated otherwise.
Absence of blinding is likely to have low risk of bias for peri-operative and efficacy
outcomes.
Free of other bias: default is low risk unless there is a fundamental flaw with the
study (e.g. inadequate follow up time for dysfunction outcomes, data not presented
for learning curve effects if these are likely to influence outcomes).
Judging overall direction of bias for individual outcomes: If confounding is judged
unbalanced, outcome should be judged as high risk of bias.
20
Appendix 2
Bibliographic details of the included studies
Al-Shaiji 2010
Al-Shaiji TF, Kanaroglou N, Thom A, Prowse C, Comondore V, Orovan W et al. A cost-analysis
comparison of laparoscopic radical prostatectomy versus open radical prostatectomy: the McMaster
Institute of Urology experience. Can Urol Assoc J 2010;4:237-41.
Anastasiadis 2003
Anastasiadis AG, Salomon L, Katz R, Hoznek A, Chopin D, Abbou CC. Radical retropubic versus
laparoscopic prostatectomy: a prospective comparison of functional outcome. Urology 2003;62:292-7.
Artibani 2003
Artibani W, Grosso G, Novara G, Pecoraro G, Sidoti O, Sarti A et al. Is laparoscopic radical
prostatectomy better than traditional retropubic radical prostatectomy? An analysis of peri-operative
morbidity in two contemporary series in Italy. Eur Urol 2003;44:401-6.
Ball 2006
Ball AJ, Gambill B, Fabrizio MD, Davis JW, Given RW, Lynch DF et al. Prospective longitudinal
comparative study of early health-related quality-of-life outcomes in patients undergoing surgical
treatment for localized prostate cancer: a short-term evaluation of five approaches from a single
institution. J Endourol 2006;20:723-31.
Barocas 2010
Barocas DA, J.A. Robotic Assisted Laparoscopic Prostatectomy Versus Radical Retropubic
Prostatectomy for Clinically Localized Prostate Cancer: Comparison of Short-Term Biochemical
Recurrence-Free Survival. J Urol 2010;183:990-6.
Chan 2008
Chan RC, Barocas DA, Chang SS, Herrell SD, Clark PE, Baumgartner R et al. Effect of a large
prostate gland on open and robotically assisted laparoscopic radical prostatectomy. BJU Int
2008;101:1140-4.
Bhayani 2003
Bhayani SB, Pavlovich CP, Hsu TS, Sullivan W, Su LM. Prospective comparison of short-term
convalescence: laparoscopic radical prostatectomy versus open radical retropubic prostatectomy.
Urology 2003;61:612-6.
Bolenz 2010
Bolenz C, Gupta A, Hotze T, Ho R, Cadeddu JA, Roehrborn CG et al. The influence of body mass
index on the cost of radical prostatectomy for prostate cancer. BJU Int 2010;106:1188-93
Brown 2004
21
Brown JA, Garlitz C, Gomella LG, McGinnis DE, Diamond SM, Strup SE. Perioperative morbidity of
laparoscopic radical prostatectomy compared with open radical retropubic prostatectomy. Urol Oncol
2004;22:102-6.
Carlsson 2010
Carlsson S, Nilsson AE, Schumacher MC, Jonsson MN, Volz DS, Steineck G et al. Surgery-related
Complications in 1253 Robot-assisted and 485 Open Retropubic Radical Prostatectomies at the
Karolinska University Hospital, Sweden. Urology 2010;75:1092-7.
Dahl 2006
Dahl DM, He W, Lazarus R, McDougal WS, Wu CL. Pathologic outcome of laparoscopic and open
radical prostatectomy. Urology 2006;68:1253-6.
Dahl 2009
Dahl DM, Barry MJ, McGovern FJ, Chang Y, Walker-Corkery E, McDougal WS. A prospective
study of symptom distress and return to baseline function after open versus laparoscopic radical
prostatectomy. J Urol 2009;182:956-65
Doumerc 2010
Doumerc N, Yuen C, Savdie R, Rahman MB, Rasiah KK, Pe BR et al. Should experienced open
prostatic surgeons convert to robotic surgery? The real learning curve for one surgeon over 3 years.
BJU Int 2010;106:378-84.
Drouin 2009
Drouin SJ, Vaessen C, Hupertan V, Comperat E, Misrai V, Haertig A et al. Comparison of mid-term
carcinologic control obtained after open, laparoscopic, and robot-assisted radical prostatectomy for
localized prostate cancer. World J Urol 2009;27:599-605.
Ficarra 2009
Ficarra V, Novara G, Fracalanza S, D'Elia C, Secco S, Iafrate M et al. A prospective, non-randomized
trial comparing robot-assisted laparoscopic and retropubic radical prostatectomy in one European
institution. BJU Int 2009;104:534-9.
Fracalanza 2008
Fracalanza S, Ficarra V, Cavalleri S, Galfano A, Novara G, Mangano A et al. Is robotically assisted
laparoscopic radical prostatectomy less invasive than retropubic radical prostatectomy? Results from a
prospective, unrandomized, comparative study. BJU Int 2008;101:1145-9.
Ghavamian 2006
Ghavamian R, Knoll A, Boczko J, Melman A. Comparison of operative and functional outcomes of
laparoscopic radical prostatectomy and radical retropubic prostatectomy: single surgeon experience.
Urology 2006;67:1241-6.
22
Greco 2010
Greco F, Wagner S, Hoda M, Kawan F, Inferrera A, Lupo A et al. Laparoscopic vs open retropubic
intrafascial nerve-sparing radical prostatectomy: surgical and functional outcomes in 300 patients.
BJU Int 2010;106:543-7.
Guazzoni 2006
Guazzoni G, Cestari A, Naspro R, Riva M, Centemero A, Zanoni M et al. Intra- and peri-operative
outcomes comparing radical retropubic and laparoscopic radical prostatectomy: results from a
prospective, randomised, single-surgeon study. Eur Urol 2006;50:98-104.
Hu 2006
Hu JC, Nelson RA, Wilson TG, Kawachi MH, Ramin SA, Lau C et al. Perioperative complications of
laparoscopic and robotic assisted laparoscopic radical prostatectomy. J Urol 2006;175:541-6.
Jacobsen 2007
Jacobsen NE, Moore KN, Estey E, Voaklander D. Open versus laparoscopic radical prostatectomy: a
prospective comparison of postoperative urinary incontinence rates. J Urol 2007;177:615-9.
Joseph 2005
Joseph JV, Vicente I, Madeb R, Erturk E, Patel HR. Robot-assisted vs pure laparoscopic radical
prostatectomy: are there any differences? BJU Int 2005;96:39-42.
Jurczok 2007
Jurczok A, Zacharias M, Wagner S, Hamza A, Fornara P. Prospective non-randomized evaluation of
four mediators of the systemic response after extraperitoneal laparoscopic and open retropubic radical
prostatectomy. BJU Int 2007;99:1461-6.
Krambeck 2009
Krambeck AE, DiMarco DS, Rangel LJ, Bergstralh EJ, Myers RP, Blute ML et al. Radical
prostatectomy for prostatic adenocarcinoma: a matched comparison of open retropubic and robotassisted techniques. BJU Int 2009;103:448-53.
Loeb 2010
Loeb S, Epstein JI, Ross AE, Schultz L, Humphreys EB, Jarow JP. Benign prostate glands at the
bladder neck margin in robotic vs open radical prostatectomy. BJU Int 2010;105:1446-9.
Malcolm 2010
Malcolm JB, Fabrizio MD, Barone BB, Given RW, Lance RS, Lynch DF et al. Quality of life after
open or robotic prostatectomy, cryoablation or brachytherapy for localized prostate cancer. J Urol
2010;183:1822-8.
23
Martorana 2004
Martorana G, Manferrari F, Bertaccini A, Malizia M, Palmieri F, Severini E et al. Laparoscopic
radical prostatectomy: oncological evaluation in the early phase of the learning curve comparing to
retropubic approach. Arch Ital Urol Androl 2004;76:1-5.
Menon 2002
Menon M, Tewari A, Baize B, Guillonneau B, Vallancien G. Prospective comparison of radical
retropubic prostatectomy and robot-assisted anatomic prostatectomy: the Vattikuti Urology Institute
experience. Urology 2002;60:864-8.
Miller 2007
Miller J, Smith A, Kouba E, Wallen E, Pruthi RS. Prospective evaluation of short-term impact and
recovery of health related quality of life in men undergoing robotic assisted laparoscopic radical
prostatectomy versus open radical prostatectomy. J Urol 2007;178:854-8.
Nadler 2010
Nadler RB, Casey JT, Zhao LC, Navai N, Smith ZL, Zhumkhawala A et al. Is the transition from
open to robotic prostatectomy fair to your patients? A single-surgeon comparison with 2-year followup. J Robotic Surg 2010;3:201-7.
Namiki 2005
Namiki S, Egawa S, Baba S, Terachi T, Usui Y, Terai A et al. Recovery of quality of life in year after
laparoscopic or retropubic radical prostatectomy: a multi-institutional longitudinal study. Urology
2005;65:517-23.
Namiki 2006
Namiki S, Egawa S, Terachi T, Matsubara A, Igawa M, Terai A et al. Changes in quality of life in
first year after radical prostatectomy by retropubic, laparoscopic, and perineal approach: Multiinstitutional longitudinal study in Japan. Urology 2006;67:321-7.
Ou 2009
Ou YC, Yang CR, Wang J, Cheng CL, Patel VR. Comparison of robotic-assisted versus retropubic
radical prostatectomy performed by a single surgeon. Anticancer Res 2009;29:1637-42.
Poulakis 2007
Poulakis V, Witzsch U, de Vries R, Dillenburg W, Becht E. Laparoscopic radical prostatectomy in
men older than 70 years of age with localized prostate cancer: comparison of morbidity,
reconvalescence, and short-term clinical outcomes between younger and older men. Eur Urol
2007;51:1341-8
Remzi 2005
24
Remzi M, Klingler HC, Tinzl MV, Fong YK, Lodde M, Kiss B et al. Morbidity of laparoscopic
extraperitoneal versus transperitoneal radical prostatectomy verus open retropubic radical
prostatectomy. Eur Urol 2005;48:83-9.
Rocco 2009
Rocco B. Robotic vs open prostatectomy in a laparoscopically naive centre: A matched-pair analysis.
BJU Int 2009;104:991-5.
Rozet 2007
Rozet F, Jaffe J, Braud G, Harmon J, Cathelineau X, Barret E et al. A direct comparison of robotic
assisted versus pure laparoscopic radical prostatectomy: a single institution experience. J Urol
2007;178:478-82.
Salomon 2002
Salomon L, Levrel O, Anastasiadis AG, Saint F, de la TA, Cicco A et al. Outcome and complications
of radical prostatectomy in patients with PSA <10 ng/ml: comparison between the retropubic, perineal
and laparoscopic approach. Prostate Cancer Prostatic Dis 2002;5:285-90.
Schroeck 2008
Schroeck FR, Sun L, Freedland SJ, Albala DM, Mouraviev V, Polascik TJ et al. Comparison of
prostate-specific antigen recurrence-free survival in a contemporary cohort of patients undergoing
either radical retropubic or robot-assisted laparoscopic radical prostatectomy. BJU Int 2008;102:2832.
Silva 2007
Silva E, Ferreira U, Silva GD, Mariano MB, Netto NR, Jr., Billis A et al. Surgical margins in radical
prostatectomy: a comparison between retropubic and laparoscopic surgery. Int Urol Nephrol
2007;39:865-9.
Soderdahl 2005
Soderdahl DW, Davis JW, Schellhammer PF, Given RW, Lynch DF, Shaves M et al. Prospective
longitudinal comparative study of health-related quality of life in patients undergoing invasive
treatments for localized prostate cancer. J Endourol 2005;19 :318-26.
Terakawa 2008
Terakawa T, Miyake H, Tanaka K, Takenaka A, Inoue TA, Fujisawa M. Surgical margin status of
open versus laparoscopic radical prostatectomy specimens. Int J Urol 2008;15:704-7.
Tewari 2003
Tewari A, Srivasatava A, Menon M. A prospective comparison of radical retropubic and robotassisted prostatectomy: Experience in one institution. BJU Int 2003;92:205-10.Touijer 2007
25
Touijer K, Kuroiwa K, Eastham JA, Vickers A, Reuter VE, Scardino PT et al. Risk-adjusted analysis
of positive surgical margins following laparoscopic and retropubic radical prostatectomy. Eur Urol
2007;52:1090-6.
Trabulsi 2008
Trabulsi EJ, Linden RA, Gomella LG, McGinnis DE, Strup SE, Lallas CD. The addition of robotic
surgery to an established laparoscopic radical prostatectomy program: effect on positive surgical
margins. Can J Urol 2008;15:3994-9.
Truesdale 2010
Truesdale MD, Lee DJ, Cheetham PJ, Hruby GW, Turk AT, Badani KK. Assessment of Lymph Node
Yield After Pelvic Lymph Node Dissection in Men with Prostate Cancer: A Comparison Between
Robot-Assisted Radical Prostatectomy and Open Radical Prostatectomy in the Modern Era. J
Endourol 2010;24:1055-60.
Wagner 2007
Wagner AA, Link RE, Trock BJ, Sullivan W, Pavlovich CP. Comparison of open and laparoscopic
radical prostatectomy outcomes from a surgeon's early experience. Urology 2007;70:667-71.
White 2009
White MA, De Haan AP, Stephens DD, Maatman TK, Maatman TJ. Comparative analysis of surgical
margins between radical retropubic prostatectomy and RALP: are patients sacrificed during initiation
of robotics program? Urology 2009;73:567-71.
26
Appendix 3 Characteristics of the included studies
Table i
Characteristics of the included studies: robotic versus laparoscopic versus open prostatectomy (n=3)
Study details
Participant characteristics
Intervention characteristics
Outcomes
Author, year: Ball 2006
Inclusion criteria: Patients with newly diagnosed
clinically localised prostate cancer.
A. Robotic prostatectomy:
Efficacy
Trade name of robot: da Vinci
PT stage
B. Laparoscopic prostatectomy: used
a well-described technique, reference
given.
Dysfunction
Language: English
Publication type: full-text
Exclusion criteria: not reported
Number of study centres: 1
A
B
C
Setting: hospital
Patient, n
urinary incontinence, erectile
dysfunction
Country: US
Recruitment/treatment dates:
January 2000-April 2005
Prospective/retrospective data
collection: prospective
Patient recruited consecutively,
Y/N: not reported
Enrolled
82
124
135
1-month
76
93
82
3-month
56
102
122
6-month
22
112
91
Age, year,
mean (SD)
60 (7)
61 (7)
59 (6)
PSA, ng/ml,
mean (SD)
6.0 (2.4) 7.2 (7.1)
C. Open prostatectomy: used a
standard radical retropubic technique.
Nerve sparing for erectile function: n
(%)
A
B
C
Length of follow up: 6 months
Source of funding: not reported
7.8 (5.6)
Clinical
stage, n (%)
T1
66 (81%) 100 (81%)
116 (86%)
Non-nerve
sparing
18
67
(22%) (54%)
40
(30%)
Unilateral
9
23
(11%) (19%)
30
(22%)
Bilateral
54
34
(66%) (27%)
65
(48%)
T2
27
Study details
Participant characteristics
T3
Intervention characteristics
15 (18%) 24 (19%)
19 (14%)
1 (1%)
0
0
Unknown
1 (1%) 0
Outcomes
0
Biopsy
Gleason
scores, n (%)
≤6
Author, year: Bolenz 2010
7
59 (72%) 94 (76%)
85(63%)
8-10
15(18%) 22 (18%)
37 (27%)
8(10%)
13 (10%)
8 (6%)
Inclusion criteria/ Exclusion criteria: not reported
Language: English
Publication type: full text paper
A
B
C
Number of study centres: 1
Patient, n
262
211
156
Setting; not reported
Age, yrs
median
62
59
61
62 (5666)
59 (5463)
61.5 (5766)
60 (5765)
56.5 (5263)
60.5 (5464)
Country: USA
Recruitment/treatment dates:
September 2003 to April 2008
BMI <30
Prospective/retrospective data
collection : not reported
BMI >30
Patient recruited consecutively,
Y/N : not reported
PSA,
ng/ml,
median
A. Robotic prostatectomy:
Safety
Nerve sparing
Blood transfusion
B. Laparoscopic prostatectomy:
Nerve sparing
C. Open prostatectomy: Nerve
sparing
28
Study details
Length of follow ups: not
reported
Participant characteristics
Intervention characteristics
Outcomes
A. Robotic prostatectomy
Safety
Robot trade name: da Vinci system;
surgical complications, op conversion,
op time, catheterisation, blood loss
(range)
BMI <30
Source of funding : not reported
BMI >30
5.2 (4.17)
5 (4.26.5)
5.6 (4.47.2)
5.4 (4.37)
5.1 (47.2)
4.7 (4.15.9)
Biopsy Gleason scores for total sample
Author, year: Drouin 2009
<6
341
7
236
8-10
48
Inclusion criteria: patients treated for prostate
cancer with surgery
Language: English
Publication type: full-text
Number of study centres: not
reported
Approaches: trans-peritoneal;
Exclusion criteria: evidence of lymph node
involvement during pre-op work-up or in case of
clinical signs of non-localised disease.
34/71 had lymph node dissection.
Efficacy
Setting: hospital
margins, PT stage, PSA recurrence
B. Laparoscopic prostatectomy
Country: France
Recruitment/treatment dates:
January 2000 – August 2004
Prospective/retrospective data
collection: retrospective
Patient recruited consecutively,
Y/N: not reported
Length of follow up: months,
mean (range), total: 49.7 (18-
A
B
C
Patient, n
71
85
83
Age, mean
(range)
60.4 (46- 61.8 (39- 60.5 (45-81)
70)
73)
BMI, mean
(range)
22.6 (22- 23 (2225)
25.2)
PSA, ng/ml, 7.8 (3mean (range) 24)
8.9 (3.437)
23.3 (22.624.8)
Approaches: trans-peritoneal;
Death
42/85 had lymph node dissection.
C. Open prostatectomy
58/83 had lymph node dissection.
9.2 (1.2-60)
29
Study details
Participant characteristics
103); A: 40.9 (18-60); B: 48.4
(18-84); C: 57.7 (18-103)
Clinical stage,
n (%)
Source of funding: not reported
T1a-b
0
0
Intervention characteristics
Outcomes
2 (2%)
T1c
50 (70%) 55(65%)
38 (46%)
17 (24%) 22(26%)
28 (34%)
4 (6%)
15 (18%)
T2a-b
T2c
8 (9%)
Biopsy
Gleason
score, n(%)
≤6
7
60 (84%) 62(73%)
59 (71%)
8-10
11(16%) 21(25%)
24 (29%)
0
0
2 (2%)
30
Table ii
Characteristics of the included studies: robotic versus laparoscopic prostatectomy (n=8)
Study details
Participant characteristics
Intervention characteristics
Outcomes
Author, year: Hu 2006
Inclusion criteria: patients had radical
prostatectomies with laparoscopic or robotic
procedures.
A. Robotic prostatectomy
Safety
Trade name of robot: da Vinci system;
Surgical complications, operation time
Language: English
Publication type: full-text
Number of study centres: 1
Setting: hospital
Approaches: trans-peritoneal;
Exclusion criteria: patients with neoadjuvant
hormonal therapy.
Country: US
Recruitment/treatment
dates:
A: June2003 – June 2004
B: October 2000 – January
2003
A
B
Patient enrolled, 671
n
517
Patient analysed, 322
n
358
62.1 (41-84) 63.7 (40-83)
BMI, median
(range)
27.5 (17.851.5)
27.4 (17.943.8)
Length of follow up: not
reported
Previous
abdominal
surgery
37/322
(11.5%)
39/358
(10.9%)
Source of funding: not
reported
PSA, ng/ml
0-4
66 (20.6%)
55 (15.4%)
4-10
213 (66.4%) 247 (69%)
Patient recruited
consecutively, Y/N: no
B. Laparoscopic prostatectomy
Approaches: trans-peritoneal
Learning curve
(both Montsouris technique)
Operating time
Nerve sparing, n (%)
A
Age, mean
(range)
Prospective/retrospective
data collection: mixture
Death
B
23
Unilateral 27
(8.4%) (6.4%)
237
Bilateral 259
(80.4%) (66.2%)
Nonsparing
35
87
(0.9%) (24.3%)
All patients (A&B) had bilateral pelvic
lymph node dissection.
31
Study details
Participant characteristics
10
42 (13.1%)
56 (15.6%)
1 (0.3%)
6 (1.7%)
0
2 (0.6%)
Intervention characteristics
Outcomes
A. Robotic prostatectomy
Dysfunction
Clinical stage, n
(%)
T1a
T1b
T1c
231 (74.5%) 261 (72.9%)
T2a
59 (19.0%)
72 (20.2%)
11 (3.5%)
4 (1.1%)
7 (2.3%)
10 (2.8%)
1 (0.3%)
1 (0.3%)
0
2 (0.6%)
5 (1.6%)
9 (2.5%)
T2b
T2c
T3a
T3b
Biopsy Gleason
score
1-5
6-7
289 (93.5%) 322 (90.2%)
8-10
15 (4.9%)
Author, year: Joseph 2005
Language:
English
Publication type:
full-text
26 (7.3%)
Inclusion criteria: last 50 patients in a series with
localised prostate cancer who had laparoscopic
radical prostatectomy or robot-assisted
prostatectomy.
urinary incontinence, erectile dysfuntion,
potency
B. Laparoscopic prostatectomy
32
Study details
Participant characteristics
Intervention characteristics
Number of study centres: 1
Exclusion criteria: first 50 cases in each
laparoscopic and robot-assisted series.
Setting: hospital
Nerve sparing, n (%)
Country: USA
Recruitment/treatment
dates: not reported
Prospective/retrospective
data collection:
retrospective
Patient recruited
consecutively, Y/N: not
reported
Length of follow up: not
reported
Source of funding: not
reported
Outcomes
A
B
Patients, n
50
50
Age, mean
(95% CI)
59.6 (1.6)
61.8 (1.6)
PSA, ng/ml,
mean (95%CI)
7.3 (1.2)
6.0 (0.83)
43(86%)
34(68%)
6 (12%)
14 (28%)
1(2%)
2 (4%)
Biopsy
Gleason score,
mean (95%CI)
6 (0.15)
6 (0.14)
Prostate size,
g, mean
(95%CI)
53 (5.3)
51 (4.1)
A
B
Unilateral
1 (2)
10 (20)
Bilateral
46 (92)
24 (48)
Non-sparing
3(6)
16 (32)
Clinical stage,
n (%)
T1c
T2a
T2b
Author, year:
Menon 2002
Inclusion criteria: patients with clinically localized
prostate cancer undergoing prostatectomy; patients
A. Robotic prostatectomy:
Equipment failure
33
Study details
Participant characteristics
Intervention characteristics
Language:
medically fit to undergo surgery, weighing <250,
waist size <45 inches, BMI <35 kg/m2 ;patients with
previous abdominal surgery were included.
First 22 patients were operated using
Montsouris technique.
English
Publication type:
full-text
Safety
Later 18 patients were operated using
Vattikuti Institute technique.
Number of study centres:
one
Outcomes
A
B
50
48
B. Laparoscopic prostatectomy:
Laparoscopic was performed using
classical Montsouris technique.
surgical complications, op time,
discharge, blood loss
Setting: hospital
Patient, n
Country: France
Recruitment/treatment
dates: October 2000 –
October 2001
Prospective/retrospective
data collection: prospective
Patient recruited
consecutively, Y/N: Y
Length of follow up: mean
(SD),
A:
3 (1.3) months
B:
8.5(3.2) months
Efficacy
Patient analysed, n
40
40
Age, mean (SD)
60.7 (7.6)
62.8 (7.0)
BMI, mean (SD)
27.7(3.2)
27.7(2.5)
PSA, ng/ml, mean (SD)
5.7 (3.2)
6.9 (4.4)
Death (none)
T1c
28(70%)
26(65%)
Learning Curve (op time)
T2
12 (30%)
14(35%)
margins, PT stage, pathological Gleason
score, PSA recurrence
Clinical stage, n
Length of follow up for
functional outcomes,
mean(SD)
A: 1.5 months
B: 6.5 months
Follow up done with
34
Study details
Participant characteristics
Intervention characteristics
Outcomes
Inclusion criteria: patients underwent robotic or
laparoscopic prostatectomy
A. Robotic prostatectomy
Safety
Robot trade name: da Vinci system.
Approaches: extra-peritoneal.
Surgical complications, operating time,
catheterisation, blood loss, blood
transfusion
B. Laparoscopic prostatectomy
Efficacy
Approaches: extra-peritoneal.
Margins, pT stage, pathological Gleason
score
telephone survey by third
party
Source of funding: not
reported
Author, year: Rozet 2007
Language: English
Publication type: full-text
A
B
133
758 (operated
at the same
period)
Number of study centres: 1
Patient, n
Setting: hospital
Country: France
Recruitment/treatment
dates: May 2003-May 2005
Prospective/retrospective
data collection: not reported
Patient recruited
consecutively, Y/N: yes for
group A
Length of follow up: not
reported
Source of funding: not
reported
Patient analysed, 133
n
133 (matchpair)
Age, mean
(range)
62.0 (49-76) 62.5 (47-74)
BMI, mean
(range)
24.8 (18.835.5)
Previous
abdominal/
A
51
25.3 (19.332.7)
51
pelvic surgery
PSA, ng/ml,
mean (range)
Clinical stage, n
(%)
Nerve sparing, n (%)
B
Death
Unilateral 35 (27.8%) 30 (23.8%)
Learning curve
Bilateral
91 (72.2%) 96 (76.2%)
Operating time
Lymph node dissection, n
7.6 (0.938.0)
7.8 (3.2-19.0)
A
B
No 131 (98.5%) 130 (97.7%)
35
Study details
Participant characteristics
Intervention characteristics
T1b
Yes 2 (1.5%)
T1c
0
1 (0.8%)
T2a
76 (57.1%)
90 (67.7%)
T2b
51 (38.3%)
39 (29.3%)
T3a
6 (4.5%)
2 (1.5%)
0
1 (0.8%)
Outcomes
3 (2.3%)
Biopsy Gleason
score, mean
(range)
Author, year: Trabulsi 2008
Language: English
≤6
6.3 (4.0-9.0) 6.3 (4.0-9.0)
7
101 (76%)
93 (70%)
8-10
29 (21.8%)
37 (27.8%)
3 (2.2%)
3 (2.2%)
Inclusion criteria: men with clinically localised
prostate cancer treated with either a robotic or
laparoscopic prostatectomy.
Publication type: full-text
Country: US
Recruitment/treatment
Safety
Used da Vinci system;
op conversion, blood loss
Surgical approaches: intra-peritoneal;
Number of study centres: 1
Setting: hospital
A. Robotic prostatectomy
A
B
Number of
patients
50
190
Age, mean
57.7 (37-60)
58.6 (43-74)
Lymph nodes dissected when indicated
(in intermediate and high risk patients) 14
(28%)
Efficacy
margins, PT stage, pathological Gleason
score
B. Laparoscopic prostatectomy
36
Study details
Participant characteristics
Intervention characteristics
dates:
(range)
Surgical approaches: trans-peritoneal
A: October 2005 –August
2006
BMI, mean
(range)
28.4 (20.4-36.6) 26.8 (18.851.8)
B: March 2000 – December
2005
PSA, ng/ml,
mean (range)
5.5 (1.1-21.1)
6.5 (0.4-46)
Prospective/retrospective
data collection: retrospective
Clinical stage, n
(%)
Patient recruited
consecutively, Y/N: not
reported
T1c
41 (82%)
145 (76%)
9 (18%)
40 (21%)
Length of follow up: not
reported
Not reported
0
5 (3%)
Source of funding: not
reported
Biopsy Gleason
score, n (%)
36 (72%)
136 (72%)
8 (16%)
31 (16%)
4 (8%)
6 (3%)
2 (4%)
3 (2%)
41 (16-102)
43.3 (14-156)
Outcomes
Lymph nodes dissection: same indication
as above. 51 (27%)
T2a
≤6
3+4
4+3
>8
Prostate size, g,
mean (range)
37
Table iii
Characteristics of the included studies: robotic versus open prostatectomy (n= 17)
Study details
Participant characteristics
Intervention characteristics
Author, year: Barocas
2010
Inclusion criteria: patients undergoing radical prostatectomy A. Robotic prostatectomy:
for clinically localised prostate cancer.
Trade name of robot: da Vinci
Language: English
Publication type: full-text
Exclusion criteria: patients with prior treatment, missing
data, lymph node involvement.
Number of study centres: 1
Setting: Medical Center
A
B
Patient, n
1413
491
Age, mean (SD)
61 (7.3)
62 (7.3)
PSA, ng/ml, median
(IQR)
5.4 (4.3-7.4)
5.8 (4.6-8.4)
Outcomes
Efficacy
margins, PT stage, pathological
Gleason score, PSA recurrence
B. Open prostatectomy: was
performed by standard techniques
with small modifications described by
Walsh and Partin
Country: USA
Recruitment/treatment
dates: June 2003-January
2008
Prospective/retrospective
data collection:
retrospective
Clinical stage, n (%)
Patient recruited
consecutively, Y/N: not
reported
T1a
3 (0.21%)
3 (0.61%)
T1b
1 (0.07%)
0
Length of follow up:
median (IQR)
T1c
1086 (77.3%)
342 (69.94%)
T2a
267 (19%)
89 (18.2%)
T2b
37 (2.63%)
42 (8.59%)
T2c
4 (0.28%)
12 (2.45%)
T3a
7 (0.5%)
0
In total: 10 (2-23) months
A: 8 (2-20) months
B: 17 (8-34) months
Source of funding: not
38
reported
T3b
0
1 (0.2%)
Missing
8 patients were missing clinical stage
2 patients were missing procedure
type
Biopsy Gleason
scores, n (%)
≤6
986 (69.8%)
327 (66.6%)
353 (25.0%)
116 (23.5%)
72 (5.1%)
48 (9.8%)
2 (0.1%)
0
7
8-10
Missing
Author, year: Carlsson
2010
Inclusion criteria: patients underwent robotic or retropubic
prostatectomy for clinically localised prostate cancer.
A. Robotic prostatectomy:
Safety
surgical complications
Language: English
Publication type:full-text
Number of study centres: 1
N
A
B
1253
485
B. Open prostatectomy:
modification of Walsh “anatomical
radical retropubic prostectomy”
Further Treatment
Urinary incontinence
Setting: hospital
Country: Sweden
Recruitment/treatment
dates: 01/2002-08/2007
Prospective/retrospective
data collection: prospective
Age, years, median
(range)
62 (35-78)
PSA, ng/ml, median
(range)
6.0 (4-9)
6.0 (4-10)
770 (61.5%)
251 (51.8%)
63 (47-77)
Both A and B: a limited lymph node
dissection preformed if indicated
(Gleason score 4+4=8 or PSA >20
ng/mL)
Death
Clinical stage, n (%)
cT1
Patient recruited
39
consecutively, Y/N: yes
cT2
435 (34.7%)
183 (37.8%)
Length of follow up:
median, A, 19 months; B, 30
months
cT3
48 (3.8%)
50 (10.4%)
Source of funding:
Swedish Cancer Society,
ALF, and the Johanna
Hagstrand & Sigfrid Linner
Foundation
Author, year: Chan
2008Language: English
Biopsy Gleason
6.3 (0.4-50)
scores, median (range)
7.4 (0.1-135)
Prostate size, ml,
median (range)
38.0 (16-130)
38.0 (16-206)
Inclusion criteria: patient with clinically localized
carcinoma of the prostate
A. Robotic prostatectomy: n= 660
Safety
Performed using a five-port technique
op conversion, op time, hospital stay
Publication type: full-text
Nerve sparing:
Number of study centres:
one
Setting: hospital
Data reported based on prostate size (large vs small). Here
we have extracted the combined data wherever possible.
When mean (range) were reported, only ranges have been
extracted.
Learning Curve
Bilateral: 86/522
Op time
Non-nerve sparing: 25/110
Country: USA
Recruitment/treatment
dates: 05/2003-08/2006
Prospective/retrospective
data collection: not
reported
Unilateral: 8/28
A
B
Patient, n
660
340
B. Open prostatectomy: n= 340
Age, range
36-78
40-81
Performed via an infra-umbilical
midline incision
PSA, ng/ml, range
0.18-76
0.5-51.7
Patient recruited
consecutively, Y/N: Y
Clinical stage, n (%)
Length of follow up: none
T1
497 (75%)
225(66%)
Source of funding: not
reported
T2
160 (24%)
111(33%)
Nerve sparing:
Unilateral: 12/30
Bilateral: 52/183
Non-nerve sparing: 52/127
40
T3
3 (1%)
4(1%)
459(70%)
212 (62%)
173 (26%)
87 (26%)
28(4%)
41(12%)
15-181
0.7-224
Biopsy Gleason scores, n
(%)
≤6
7
8-10
Prostate size, g, range
Author, year: Doumerc
2010
Inclusion criteria: Clinically localised prostate cancer
Language: English
Publication type: Full-text
paper
Number of study centres:
Not reported
Exclusion criteria: Patients with factors considered to
increase surgical difficulty eg. Morbid obesity, prostate size
>100ml, large middle lobe, previous TURP, a history of Lap
hernia mesh repair, multiple abdominal operation, high
volume tumer
A. Robotic prostatectomy: Describe
by Patel. Trans peritoneal surgical
approach, Trade name and
manufacturer of robot not reported
B. Open prostatectomy: Via infra
umbilical incision
Safety
surgical complications, op time,
hospital stay, catheterisation, blood
loss
Efficacy
margins,PT stage, pathological
Gleason score
Setting: Referral Institution
A
B
Patient, n
212
502
Age, mean
(range)
61.3 (41-76)
60.1 (40-78)
PAS level
7.1 (0.7-41)
8.3 (0.9-64)
Lymph node dissection, n:
Country: Australia
Recruitment/treatment
dates: February 2006December 2008
Prospective/retrospective
data collection:
Prospective
Patient recruited
A
B
No
lymph
node
158 /212
(74.5%)
239/502
(47.6%)
Negative
54/54
(100%)
247/263
(94%)
1 positive
0
11/263
Death
Clinical Stage, n
(%)
41
consecutively, Y/N: Yes
T1a
4 (2%)
5 (1%)
Length of follow up:
Months A:11.2 (9.4)
T1b
2 (1%)
5 (1%)
T1c
99 (47%)
201 (40%)
T2a
59 (28%)
111 (22%)
T2b
16 (7%)
70 (14%)
T2c
32 (15%)
95 (19%)
T3
0
15 (3%)
73 (34%)
126 (25%)
128 (61%)
321 (64%)
12 (5.6%)
55 (11%)
50 (16-140)
53.2 (20-145)
(4%)
>1
positive
0
5/263
(2%)
B: 17.2 (9.7)
Source of funding: NIH
Grant 5R01DK077116
Gleason Score, n
(%)
≤6
7
8-10
Prostate size ml
Author, year: Ficarra 2009
Inclusion criteria: all patients undergoing robotic or open
prostatectomy for clinically localised prostate cancer
Language: English
Publication type: full-text
A. Robotic prostatectomy
Safety
Trade name of robot: da Vinc system;
surgical complications, op time,
hospital stay, catherisation, blood loss
Approaches: extra-peritoneal;
Exclusion criteria: not reported.
Number of study centres: 1
64 (62%) had bilateral nerve sparing;
A
Efficacy
B
Setting: hospital
Patient, n
103
105
Age, median (IQR)
61 (57-67)
65 (61-69)
Lymph node dissected in patients with
high risk of lymph node involvement.
margins, PT stage
B. Open prostatectomy
Dysfunction
Country: Italy
Recruitment/treatment
dates: February 2006-April
42
2007
BMI, median (IQR)
26 (24-28)
26 (24-28)
Approaches: extra-peritoneal;
Prospective/retrospective
data collection: prospective
PSA, ng/ml, median
(IQR)
6.4 (4.6-9)
6 (5-10)
41 (39%) had bilateral nerve sparing;
Patient recruited
consecutively, Y/N: yes
Clinical stage, n (%)
urinary incontinence, erectile
dysfunction
Same indication as above for lymph
node dissection.
T1c
77 (75%)
66 (63%)
T2a-b
22 (21%)
32 (30%)
T2c
4 (4%)
7 (7%)
Length of follow up: 1 year
Source of funding: partially
funded by the Italian
Ministry for University and
Research
Biopsy Gleason score, N=97
n (%)
N=104
<6
71 (73%)
67 (64%)
18 (19%)
29 (28%)
8 (8%)
8 (8%)
37.5 (30-48)
40 (30-47)
7
8-10
Prostate size, ml,
median (IQR)
Author, year:
2008
Fracalanza
Inclusion criteria: patients with clinically localized
prostate cancer (cT1-2)
A. Robotic prostatectomy:
Safety
Trade name: da Vinci
surgical complications, op time,
hospital stay, blood loss, surgical
incision, time to mobilisation, oral
feeding
Language: English
Publication type: full-text
Number of study centres:
1
A
B
Patient , n
35
26
Age, mean (range)
62
68.5
Setting: hospital
(56-68)
(59-71)
performed with transperitoneal
approach with an antegrade prostatic
dissection
Efficacy
Lymph node dissection was carried
out in the men with a high risk of
lymph node involvement
margins, PT stage
43
Country: Italy
BMI, kg/m2, mean (SD) 25.5(2.7)
26.4(3.7)
Recruitment/treatment
dates: May 2006- October
2006
PSA, ng/ml, median
(range)
6.2
6.2
B. Open prostatectomy:
Learning curve
(4.2- 10.2)
(4.5-9.1)
Performed according to the Walsh
technique
Op time
Prospective/retrospective
data collection: prospective
Biopsy Gleason score,
n (%)
14(40%)
6(23%)
Patient recruited
consecutively, Y/N: Y
≤6
13(37%)
16(62%)
8(23%)
4(15%)
40
36
(30-60)
(30-40)
4(3-4)
4.5(3.7-5)
7
Length of follow up: None
All patients had lymph node
dissection, including external iliac and
obturatory lymph nodes.
8-9
Source of funding: The
Italian ministry for
University and Research
Prostate size, ml,
median (range)
Charlson score, mean
(SD)
Author, year: Krambeck
2009
Inclusion criteria: patients undergoing clinically localised
prostate cancer.
A. Robotic prostatectomy:
Safety
Robot trade name: da Vinchi system.
surgical complications, Op time,
hospital stay
Language: English
Publication type: full-text
All patients had pelvic
lymphadenectomy.
Exclusion criteria: not reported.
Number of study centres: 1
A
B
Efficacy
Setting: clinic
Patient , n
294
588
B. Open prostatectomy:
Country: US
Age, median (range)
61.0 (38.0-76.0)
61.0 (41.0-77.0)
All patients had pelvic
lymphadenectomy.
Recruitment/treatment
dates: August 2002December 2005
PSA, ng/ml, median
(range)
4.9 (0.5-33.5)
5.0 (0.6-39.7)
margins, Path Gleason, PSA
recurrence, local recurrence,
metastatic recurrence
Dysfunction
Nerve-sparing, n
44
Prospective/retrospective
data collection:
retrospective
Clinical stage, n (%)
T1a/b
0
4 (0.7%)
Patient recruited
consecutively, Y/N: yes in
the robotic group; no in the
open group.
T1c
214 (72.8%)
418 (71.1%)
T2a
75 (25.5%)
130 (22.1%)
T2b
4 (1.4%)
28 (4.8%)
Length of follow up:
median 1.3 years
T3/4
1 (0.3%)
8 (1.4%)
Source of funding: not
reported
Biopsy Gleason score,
n (%)
A
B
Unilateral
20 (6.8%)
26 (4.4%)
Bilateral
221 (75.1%) 509
(86.6%)
urinary incontinence, erectile
dysfunction
Death
Learning curve
op time
≤6
214(72.8%)
441 (75.0%)
70 (23.8%)
133 (22.6%)
10 (3.4%)
14 (2.3%)
7
8-9
Author, year: Loeb 2010
A. Robotic prostatectomy: various
techniques but the prostatic dissection
was always ante grade, with division
of the BN from anterior and posterior
Inclusion criteria: not reported
Language: English
Publication type: Full-text
paper
Number of study centres:
Not reported
Exclusion criteria: not reported
A
B
Total
No. of
Patients
152
137
289
Age,
mean
(SD)
NR
NR
58.1 (5.6)
PSA
level
NR
NR
5.4 (2.9)
Setting: Medical Institution
Country: USA
Recruitment/treatment
dates: 2005-2008
Prospective/retrospective
Efficacy
margins, PSA recurrence
B. Open prostatectomy: performed
in the standard anatomical fashion
described by Latiff and Gomez
45
data collection:
Prospective
mean
(SD)
Patient recruited
consecutively, Y/N: Not
reported
Clinical
Stage, n
(%)
Length of follow up: Not
reported
T1c
Source of funding: Not
reported
NR
NR
T2
220 (76.1)
T3
67 (23.1)
Missing
1 (0.4)
1 (0.4)
Gleason
score, n
(%)
≤6
7
8-10
NR
NR
199 (68.9)
73 (25.2)
17 (5.9)
Author, year: Malcolm
2010
Language: English
Inclusion criteria: undergoing operative treatment for
localised prostate cancer. Included in the analysis if a
baseline and at least one follow up questionnaire were
completed (149 excluded)
A. Robotic prostatectomy: Nerve
sparing techniques used where
clinically appropriate as determined
by the surgeon
Dysfunction
urinary function, sexual function
Publication type: Full-text
paper
Number of study centres:
Single centre
Exclusion criteria: patients were excluded from the analysis B. Open prostatectomy: Nerve
if multimodal treatment was administered. 195 patients with sparing techniques used where
a UCLA-PCI function/bother score <30 at baseline excluded clinically appropriate as determined
46
Setting: Prostate
centre/Institution
by the surgeon. Retropubic or perineal
route
from stat analysis
Country: USA
Recruitment/treatment
dates: February 2000December 2008
Prospective/retrospective
data collection:
Prospective
A
B
No of patients
447
135
Age, mean (SD)
59 (6)
59 (7)
Nerve sparing, n (%)
T1c or less
Length of follow up:
T2b
A: 20 months
Unknown
Source of funding: Not
reported
3 authors declared financial
interest with In Touch
Health Inc, Endocare Inc,
Intuitive Surgical, Dendreon
Crop, southwest Oncology
Group, ContraVac and
Theralogix
Author, year: Miller 2007
B
Spared
366
(82%)
95 (70%)
Not
spared
81 (18%)
40 (30%)
Clinical Stage, n
(%)
Patient recruited
consecutively, Y/N: Not
reported
B: 31.5 months
A
340 (76%)
112 (83%)
68 (15%)
17 (13%)
32 (7%)
6 (4%)
7 (2%)
0
269 (60%)
93 (69%)
154 (34%)
34 (25%)
24 (5%)
8 (6%)
5.2 (3.9, 6.8)
5.7 (4.7, 7.3)
T2a
Biopsy Gleason
Score, n (%)
6 or less
7
8+
PSA, ng/ml,
median (IQR)
Inclusion criteria: patients with clinically localised (cT1-2)
prostate cancer.
A. Robotic prostatectomy
Safety
Robot trade name: da Vinci system (4
47
robotic and 2 assistant ports in a
manner similar to Menon)
Language: English
Publication type: full-text
blood loss
Exclusion criteria: not reported
Number of study centres: 1
A
Quality of life
B
Setting: hospital Institution
Patient , n
42
120
Country: US
Age, mean
61.1
60.6
B. Open prostatectomy: Anatomical
retropubic radical prostectomy via a
10 -12cm infraumbilical medline
incision
Recruitment/treatment
dates: 7/2002-8/2006
For both A & B, nerve sparing was
performed when oncologically
appropriate and in patient who were
potent pre-op.
Prospective/retrospective
data collection: prospective
Patient recruited
consecutively, Y/N: not
reported
Length of follow up: 6
weeks
Source of funding: not
reported
Author, year: Nadler 2010
A. Robotic prostatectomy: 4 arm, 5
port technique
Inclusion criteria: not reported
surgical complications, op time,
hospital stay, blood loss
Language: English
Publication type: Full-text
paper
Number of study centres:
Single centre
Safety
Exclusion criteria: not reported
A
B
No. of patients
50
50
Age, mean
(range)
59.7 (44-77)
60 (40-75)
B. Open prostatectomy: Performed
as described by Mc Carthy and
Catalona
Efficacy
margins, PT stage, PSA recurrence
Setting: Not reported
Nerve sparing, n (%)
48
Country: USA
Recruitment/treatment
dates: July 2002-Feb 2006
(B)
Oct 2005-Oct 2006 (A)
Prospective/retrospective
data collection: both
BMI mean
(range)
28.6 (23.3-42)
28.2 (21-42.6)
PSA level mean
(range)
6.5 (1.5-18.8)
8.5 (1.9-95.6)
Clinical stage, n
(%)
41 (82%)
41 (82%)
9 (18%)
9 (18%)
6.66 (6-10)
A
B
Dysfunction
Bilateral
38 (76%)
43 (86%)
urinary continence, potency
Unilateral
8 (16%)
0
Nonnerve
sparing
4 (8%)
7 (14%)
T1
Patient recruited
consecutively, Y/N: yes
T2
6.42 (6-9)
Length of follow up: 2
years
Biopsy Gleason
Score, mean
(range)
Prostate size
(ml), mean
(range)
49.4 (27.2-109.1)
Lymph node dissection, n:
A: 29/50 (58%)
B: Total 50/50 (100%)
Source of funding: Not
reported
62.8 (14.9-135.8)
AUA risk
stratification, n
A
B
Bilateral
16 (55%)
45 (90%)
Unilateral
13 (45%)
5 (10%)
Low
30 (60%)
28 (56%)
14 (28%)
12 (24%)
6 (12%)
10 (20%)
Moderate
High
Author, year: Ou 2009
Inclusion criteria: patients undergoing prostatectomy.
Language: English
Publication type: full-text
A
B
A. Robotic prostatectomy:
performed as described by Patel with
minor modification.
22/30 (73.3%) patients had bilateral
lymph nodes dissection.
Safety
op conversion, surgical
complications, op time, hospital stay,
catherisation, blood loss
49
Number of study centres: 1
Patients, n
30
30
Setting: hospital
Age, mean (SD)
67.3 (6.2)
70.0 (6.1)
Country: Taiwan
BMI, mean (SD)
24.2 (3.2)
24.1 (3.3)
B. Open prostatectomy: performed
using Walsh’s technique.
16.5 (18.8)
15.9 (14.1)
30/30 (100%) patients had bilateral
lymph nodes dissection.
T1
15 (50%)
9 (30%)
Nerve-sparing, n (%)
T2
15 (50%)
19 (63.3%)
Patient recruited
consecutively, Y/N: yes
T3
0
2 (6.7%)
Length of follow up: 15
months
Biopsy Gleason score,
mean (SD)
6.1 (0.9)
6.2 (1.6)
Recruitment/treatment
dates: 4/2004-4/2007
PSA, mean (SD)
Efficacy
margins, Path Gleason PSA
recurrence
Clinical stage, n (%)
Prospective/retrospective
data collection:
retrospective
Dysfunction
A
Unilateral
B
Urinary incontinence, erectile
dysfunction
5 (16.7%) 1
(3.3%)
Learning curve
Bilateral
11(36.7%) 1
(3.3%)
op time
Source of funding: NR
Non-nerve sparing 14
(46.7%)
Author, year: Rocco 2009
Inclusion criteria: patients had robotic or laparoscopic
prostatectomy
28
(93.3%)
A. Robotic prostatectomy: As
described by Patel technique
Safety
op time, hospital stay, catherisation,
blood loss
Language: English
Publication type: full-text
B. Open prostatectomy: Walsh
technique
Exclusion criteria:
Number of study centres: 1
Efficacy
Setting: Institution
A
B
120
240
Country: Italy
Patient, n
margins, PT stage, pathological
Gleason score
Recruitment/treatment
dates: A: November 2006-
50
December 2007
Age, median (range)
63 (47-76)
B: May 2004-February 2007
PSA, ng/ml, median (range)
6.9 (0.4-23.0) 6.7 (0.722.0)
Prospective/retrospective
data collection:
Clinical stage, n (%)
A: prospective
T1c
82 (69%)
145 (6%)
B: retrospective
T2a
36 (31%)
93 (39%)
Patient recruited
consecutively, Y/N: yes in
laparoscopic group
Biopsy Gleason score, median
(range)
6 (4-9)
6 (4-10)
Dysfunction
63 (46-77)
urinary incontinence, erectile
dysfunction
Length of follow up: 1 year
Source of funding: not
reported
Author, year: Schroeck
2008
Inclusion criteria: not reported
A. Robotic prostatectomy: Trade
name: da Vinci performed using
Vattikuti Institute technique
Safety
B. Open prostatectomy
Efficacy
blood loss
Language: English
Exclusion criteria: Conversion to open procedure
Publication type: full-text
A
B
Patient, n
362
435
Age, median(range)
59.2 (54.5-63.8)
60.3 (55.3-64.7)
Lymph node dissection, n:
BMI, median (range)
27.8 (25.7-29.9)
27.7 (25.5-30.4)
A: 271/362 (74.9%)
PSA, ng/ml, median
(range)
5.4 (4.1-7.1)
5.3 (4.1-7.2)
B: Total 313/435 (72%)
Number of study centres: 1
margins, pathological Gleason score,
PSA recurrence
Setting: not reported
Country: USA
Recruitment/treatment
dates: August 2003 to
January 2007
Prospective/retrospective
data collection:
Clinical stage, n (%)
51
retrospective
T1
281 (77.6%)
296 (68%)
Patient recruited
consecutively, Y/N: yes
T2
57 (15.8%)
101 (23%)
T3
0
12 (3%)
Not reported
24 (6.6%)
26 (6%)
254 (70 %)
241(55%)
89 (24.6%)
127 (29%)
9 (2.5%)
42 (11%)
10 (2.8%)
25 (5 %)
42.9 (34.3-55)
41.3 (24.4-52)
Length of follow up:
Robotic, mean 1.09 years;
Open, mean 1.37 years
Source of funding: not
reported
Biopsy Gleason
scores, n (%)
≤6
7
8-10
Not reported
Prostate size, ml,
median (range)
Author, year: Tewari 2003
Language: English
Inclusion criteria: patients with clinically localised prostate
cancer, patients who had a 10 years life expectancy and had
prostate cancer of gleason score ≥6
A. Robotic prostatectomy: used da
Vinci system (robotically assisted
Vattikuti Institute prostatectomy)
Publication type: full-text
Number of study centres: 1
A
B
Setting: hospital
Patient, n
200
100
Country: US
Age, mean (range)
59.9 (40-72)
63.1 (42.8-72)
Recruitment/treatment
Safety
op conversion, surgical
complications, hospital stay,
catheterisation, blood loss
B. Open prostatectomy: Conducted
using the anatomical technique
Efficacy
For A & B, some patients had lymph
margins, PT stage, Path Gleason,
PSA recurrence
52
dates: October1999December 2002
Prospective/retrospective
data collection: prospective
Patient recruited
consecutively, Y/N: yes for
open group, not reported for
robotic group.
Length of follow up: A:
mean 236 days; B: 556 days
BMI, mean (range)
27.7 (19-38)
27.6 (17-41)
node dissection
Previous abdominal & 20%
hernia surgery
19%
Dysfunction
PSA, ng/ml, mean
(range)
7.3 (1.9-35
urinary incontinence, erectile
dysfunction
6.4 (0.6-41)
Clinical stage, % as
reported by study
authors
Quality of life
Pain
T1a
T1c
0.5%
0
T2a
49%
59%
T2b
10%
10%
T3a
39%
35%
1.5%
4%
67%
52%
28%
35%
6%
13%
6.5
6.6
58.8 (18-140)
48.4 (24.2-70
Death (none)
Biopsy Gleason
scores, n
≤6
7
8-10
Mean
Prostate size, ml,
mean (range)
53
Author, year: Truesdale
2010
Language: English
Inclusion criteria: Patients who had undergone open or
robot-assisted radical prostatectomy with concurrent pelvic
lymph node dissection for histologically proven, clinically
localised prostate cancer.
A. Robotic prostatectomy:
Safety
Pelvic lymph node dissection done.
Positive lymph node 1/99 (1)
op time, blood loss
Exclusion criteria: Not reported
B. Open prostatectomy:
Publication type: Full-text
paper
Efficacy
PT stage, pathological Gleason score
Number of study centres:
Single centre
Pelvic lymph node dissection done.
Positive lymph node 19/217 (8.8)
Setting: Academic Istitution
A
B
Country: USA
Patient, n
99
217
Recruitment/treatment
dates: January 2005November 2009
Age, mean (SD)
59.2 (7.1)
61.7 (6.8)
BMI , mean (SD)
24.6 (8.3)
23.1 (9.1)
Prospective/retrospective
data collection:
Retrospective
PSA ng/ml,
mean (SD)
7.04 (7.5)
8.35 (7.62)
57 (57.6%)
155 (71.4%)
4 (4%)
12 (5.5%)
38 (38.4%)
50 (23%)
28 (28.3%)
63 (29%)
Patient recruited
consecutively, Y/N: Not
reported
Length of follow up: Not
reported
Overall lymph node positivity rate
6.3%
Clinical stage, n
(%)
T2a
T2b
T2c
Source of funding: Not
reported
Biopsy Gleason
score, n (%)
≤6
54
7
34 (34.3%)
95 (43.8%)
8-10
37(3.4%)
59 (27.2%)
43 (43.4%)
64 (29.5%)
36 (36.4%)
94 (43.3%)
20 (20.2%)
59 (27.2%)
D’Amico risk, n
(%)
Low
Intermediate
High
Author, year: White 2009
Inclusion criteria: patients had clinically localised
carcinoma of the prostate
A. Robotic prostatectomy: technique
as described by Menon et al.
Language: English
Safety
op conversion
Publication type: full-text
Number of study centres:
1
Setting: community
urologic practice
A
B
Patient, n
50
50*
Age, mean
62
64.7
PSA, ng/ml, mean
4.63
5.04
Country: US
B. Open prostatectomy: performed
in the traditional fashion.
Efficacy
margins, PT stage, pathological
Gleason score
For both A&B, nerve sparing was
performed in all patients, but not
reported unilateral or bilateral.
Clinical stage, n (%)
Recruitment/treatment
dates: December 2005March 2008
T1
40 (80%)
38 (76%)
T2
10 (20%)
12 (24%)
Prospective/retrospective
data collection:
55
retrospective ; laparoscopic
procedures were conducted
before the initiation of the
robotic programme
T3
Patient recruited
consecutively, Y/N: yes in
robotic group, no in the
laparoscopic group
≤6
Length of follow up: not
reported
Source of funding: not
reported
0
0
Biopsy Gleason scores, n
(%)
39 (78%)
40 (80%)
10 (20%)
9 (18%)
1 (2%)
1 (2%)
7
8-10
*matched to the robotic group according to clinical stage,
baseline PSA level, age, Gleason score
56
Table iv
Characteristics of the included studies: laparoscopic versus open prostatectomy (n=27)
Study details
Participant characteristics
Intervention characteristics
Outcomes
Author, year: Al-Shaiji
2010
Inclusion criteria: Those diagnosed with organconfined prostate cancer
A. Laparoscopic prostatectomy
Not reported
Safety
blood loss, operating time, hospital stay
Language: English
Publication type: full-text
Exclusion criteria: Not reported
Number of study centres:
Single
B. Open prostatectomy
A
B
Patient, n
70
70
Age, mean
(range) SD
60 (48-73)
62 (46-75)
5.84
6.33
0-10
67
56
>10
3
14
55 (78.6%)
41(58.6%)
14 (20%)
24 (34.3%)
1(1.4%)
3 (4.3%)
0
2 (2.8%)
Not reported
Setting: Health Centre
Country: Canada
Recruitment/treatment
dates: November 2004November 2005
Prospective/retrospective
data collection:
Retrospective
PSA level
Patient recruited
consecutively, Y/N: Yes
Clinical stage, n
(%)
Length of follow up: Not
Reported
T1c
Source of funding: Not
Reported
T2a
T2b
T2c
57
Biopsy gleason
score, n (%)
<7
34 (48.6%)
33 (47.1%)
32 (45.7%)
30 (42.9%)
4 (5.7%)
7 (10%)
7
>7
Author, year:
Anastasiadis 2003
Inclusion criteria: men with localised prostate
cancer
A. Laparoscopic prostatectomy:
Performed with a descending technique.
Safety
catheterisation, surgical complications
Language: English
Publication type: full-text
Number of study centres:
1
Exclusion criteria: Patients using vacuum erection
devices, pharmacologic injection therapy or
transurethral alprostadil were not included in the
questionnaire group.
Setting: Hospital
Country: France
Recruitment/treatment
dates: May 1998 to
December 2001
Prospective/retrospective
data collection:
prospective
A
B
Patient, n
230
70
Age, mean
(range) SD
64.1 (46-77) 6.4 64.8 (50-75) 6.4
PSA level, mean 10.7 (1. 2-80)
(range) SD
8.8
Patient recruited
consecutively, Y/N: yes
Clinical stage, n
(%)
Length of follow up:
laparoscopic median 15.1
months; open median 15.5
T1a-b
10 (4.3%)
B. Open prostatectomy: Performed with
an ascending technique.
Efficacy
margins, PT stage, pathological Gleason
score
For both interventions, the indication for
preserving one (laparoscopic n=33
(14.3%); open n= 4 (5.7%)) or both
bundles (laparoscopic n=77 (33.4%); open
n=28 (40%)) depended on pre and intraoperative factors. If all biopsies from one
lobe were positive that bundle was usually
sacrificed prioritising cancer control
before sexual function
Dysfunction
urinary continence
11.2 (1.2-70) 9.7
2 (2.8%)
T1c
58
months
T2a
156 (67.8%)
50 (71.4%)
Source of funding: not
reported
T2b
58 (25.2%)
17 (24.3%)
6 (2.6%)
1 (1.4%)
Biopsy gleason, 5.8 (2-9) 1.2
mean (range) SD
Author, year: Artibani
2003
6.1 (3-10) 1.1
Inclusion criteria: patients undergoing
prostatectomy
A. Laparoscopic prostatectomy
Safety
Surgical approaches: extra-peritoneal
hospital stay, catheterisation, surgical
complications
Language: English
Publication type: full-text
A
B
Patient, n
71
50
Age, mean (SD)
63 (5.8)
64 (6.6)
PSA, ng/ml, mean (SD)
15.7 (17)
11 (9)
T1b
1 (1.5%)
4 (8%)
T1c
20 (28%)
26 (52%)
T2a
34 (48%)
15 (30%)
T2b
10 (14%)
4 (8%)
T3
6 (8.5%)
1 (2%)
Biopsy Gleason score,
mean (SD)
5.8 (1.3)
5.7 (1.2)
B. Open prostatectomy
Number of study centres:
2
Setting: hospital
Efficacy
A
Country: Italy
Clinical stage, n (%)
Recruitment/treatment
dates: January 2001December 2001
Prospective/retrospective
data collection: not
reported
Patient recruited
consecutively, Y/N: yes
Length of follow up:
months, median (range), A,
10 (4-16); B, 10 (4-18)
margins, PT stage, pathological Gleason
score, PSA recurrence
Nerve sparing, n (%)
B
Unilateral
9 (12.7%) 0
Dysfunction
Bilateral
9 (12.7%) 0
urinary incontinence, erectile dysfunction
Non-sparing
53(74.6%) 50 (100%)
Lymph node dissection:
A: not done if PSA <10 ng/ml & biopsy
Gleason score <7.
B: All had lymph node dissection
59
Source of funding: not
reported
Additional information:
two groups of patients were
from two different hospitals
in the same city.
Author, year: Bhayani
2003
Language:
Inclusion criteria: All patients undergoing
laparoscopic and open radical prostatectomy for
localised prostate cancer
A. Laparoscopic prostatectomy:
Performed using the Guillonneau &
Vallancien technique
Safety
Exclusion criteria: not reported
B. Open prostatectomy: Performed using
the Walsh technique
Efficacy
English
op conversion, op time, hospital stay,
surgical complications, catheterisation,
blood loss
Publication type: full-text
Number of study centres:
1
Setting: Urological
institute/medical centre
A
B
Patients, n
33
24
Age, mean (SD)
57.4 (6.3)
60.5 (6.4)
PSA, ng/ml, mean (SD)
6.74 (3.8)
8.6 (9.1)
T1a
0
1(4.2%)
T1c
21(63.6%) 14 (58.3%)
T2a
11(33.3%) 8 (33.3%)
T2b
1 (3.1%)
Biopsy Gleason score,
mean (SD)
6.06 (0.25) 6.13 (0.44)
PT stage
Country: USA
Recruitment/treatment
dates: July 2001 to June
2002
Prospective/retrospective
data collection:
retrospective
Patient recruited
consecutively, Y/N:
unclear
Length of follow up: not
Clinical stage, n (%)
1(4.2%)
60
reported
Source of funding: not
reported
Author, year: Brown 2004
Inclusion criteria:
Language: English
Publication type: full-text
Number of study centres:1
Exclusion criteria: patients requiring conversion to
open procedure and patients receiving neo-adjuvant
hormonal therapy or with metastatic disease.
A. Laparoscopic prostatectomy:
Performed using the Guillonneau &
Vallancien technique. Simultaneous
bilateral pelvic lymph node dissection
performed in 11 patients
Safety
op time, hospital stay, re-admission,
surgical complications
Efficacy
Setting: Urologic Institution
Country: USA
Recruitment/treatment
dates: March 2000 to
March 2002
Prospective/retrospective
data collection:
prospective
A
B
Patient, n
60
60
Age, mean
(median)
58.8 (58.5)
59 (59)
PSA, ng/ml, mean 6.4 (6)
(rmedian)
Patient recruited
consecutively, Y/N: yes
Clinical stage, n
(%)
Length of follow up:
T1a-b
Source of funding: not
reported
T1c
B. Open prostatectomy: Performed in the
standard fashion with simultaneous
modified bilateral pelvic lymph node
dissection. Unilateral or bilateral nerve
sparing were performed when indicated.
margins, PT stage
Learning curve
op time
5.6 (5.1)
0
1
47 (78.3%)
45
13 (21.7%)
11
0
3
T2a
T2b
Biopsy Gleason
61
scores, n (%)
Author, year: Dahl 2009
Language: English
Publication type: full-text
Number of study centres:
1
≤6
47 (78.3%)
41 (68.3%)
7
13 (21.7%)
18 (30%)
8-10
0
1 (1.7%)
Inclusion criteria: patients 40-70 years old
scheduled to undergo open or laparoscopic radical
prostatectomy for clinical stage T1-2 N0M0 prostate
cancer by any one of 3 experienced surgeons.
A
Patient, n
6 months
75
78
Prospective/retrospective
data collection: prospective
12 months
78
73
Age, y, mean
59.5
59.9
0-2.5
12 (12%)
11 (11%)
2.6-4.0
20 (19%)
26 (25%)
4.1-7.0
42 (40%)
40 (39%)
7.1-100
17 (16%)
14 (14%)
Source of funding: not
reported
B. Open prostatectomy
B
At baseline
Length of follow up: 12
months
surgical complications
104
urinary incontinence, erectile dysfunction
Nerve-sparing, n (%)
Recruitment/treatment
dates: 16th June 2003 – 22nd
July 2004
Patient recruited
consecutively, Y/N: yes
Safety
Dysfunction
Exclusion criteria: not reported
Setting: hospital
Country: US
A. Laparoscopic prostatectomy
A
B
Unilateral
5 (5%)
4 (4%)
Further Treatment
Bilateral
98 (94%)
98 (96%)
Cancer treatment
Non-nerve
sparing
1 (1%)
0
102
PSA, ng/ml, n (%)
62
>100
Author, year: Dahl 2006
13 (13%)
11 (11%)
Inclusion criteria: patients who underwent radical
prostatectomy
Language: English
A. Laparoscopic prostatectomy: N=286
performed using modified Guillonneau &
Vallancien technique.
Efficacy
margins, PT stage, pathological Gleason
score
Publication type: full-text
Exclusion criteria: not reported
Number of study centres:
1
B. Open prostatectomy: N=714
Setting: hospital
Baseline characteristics:
Country: USA
PSA level=10ng/mL in >90% patients.
Recruitment/treatment
dates: 2001 to 2005
T1c=89%
Prospective/retrospective
data collection: not
reported
Quote as “similar distributions of clinical stages,
preoperative PSA levels and Gleason scores on
biopsy were seen between two groups.”
Patient recruited
consecutively, Y/N: yes
Length of follow up: not
reported
Source of funding: not
reported
Author, year: Ghavamian
2006
Inclusion criteria: clinically localised prostate
cancer with low comorbidities and a greater than 10year life expectancy
Language: English
Publication type: full-text
Exclusion criteria: not reported
A. Laparoscopic prostatectomy:
Performed using the Stolzenburg et al and
Bollens et al technique. Extraperitoneal,
n=40; transperitoneal n=30. Nerve sparing
performed when appropriate.
Lymphadenectomy performed when PSA
>10ng/nL or Gleason score ≥ 7.
Safety
op conversion, surgical complications, op
time, hospital stay, blood loss
Dysfunction
63
Number of study centres:1
A
B
70
70
Setting:
Patient, n
Country: USA
Age, mean (range) 60.8 (43-72)
SD
6.1
Recruitment/treatment
dates: laparoscopic, 2001 to
2001; open, 1999 to 2001
Prospective/retrospective
data collection:
retrospective
57.8 (44-72)
7.3
urinary incontinence, erectile dysfunction
B. Open prostatectomy: Performed using
modified Walsh technique. Nerve sparing
performed when appropriate.
Lymphadenectomy performed when PSA
>10ng/nL or Gleason score ≥ 7.
PSA, mean (range) 7.6 (3-16.5) 8.0 9.9 (2.3-33.7)
SD
7.1
Clinical stage, n
(%)
54 (77.1%)
49 (70%)
7 (10%)
9 (12.9%)
9 (12.9%)
12 (17.1%)
6.4 (0.8)
6.7 (1.3)
49 (70%)
43 (61.4%)
19 (27.1%)
21 (30%)
2 (2.9%)
6 (8.6%)
T1c
Patient recruited
consecutively, Y/N: unclear
Length of follow up: at
least 18 months
Source of funding: not
reported
T2a-b
T2c
Biopsy Gleason
score, mean (SD)
n
5-6
7
8-10
Prostate volume
40.8 (20-114)
cm3, mean (range)
Author, year: Greco 2010
53.2 (19-135)
Inclusion criteria: PSA level <10 ng/ml, Gleason
score ≤ 7 and only two positive of at least 12 biopsy
A. Laparoscopic prostatectomy: Nerve
sparing LRP
Safety
op conversion, surgical complications, op
64
Language: English
core
Publication type: full-text
Exclusion criteria: not reported
Number of study centres:
Single
time, catheterisation, blood loss
B. Open prostatectomy: Nerve sparing
RRP
A
B
Efficacy
Patient, n
150
150
margins, PT stage
Age, mean
(range)
60.5 (45-76)
61.5 (49-74)
Country: Italy
Recruitment/treatment
dates: Jan 2005-Nov 2007
BMI, mean
(range)
32 (26-38)
29 (25-53)
Prospective/retrospective
data collection: Prospective
PSA level, mean 6.3 (2.4-10)
(range)
Patient recruited
consecutively, Y/N: Yes
Clinical stage, n
(%)
Length of follow up: 1 year
T1a
Source of funding: Not
reported
T1b
Setting: Clinic
Dysfunction
urinary incontinence, erectile dysfunction
6.95 (3.4-10)
18 (12%)
15(10%)
23(15.3%)
20 (13.3%)
106 (70.7%)
110 (73.3%)
3 (2%)
5 (3.3%)
5 (3-7)
5 (3-7)
T1c
T2a
Biopsy gleason
score, mean
(range)
Prostate size, ml, 45 (18-72)
mean (range)
Author, year:
2006
Guazzoni
54 (20-88)
Inclusion criteria:
consecutive and age
matched patients who were diagnosed with clinically
localized prostate cancer (cT1- cT2); patients who
A. Laparoscopic prostatectomy:
performed according to Montsouris
technique; the urethra-vesicle anastomosis
Safety
op conversion, surgical complications, op
65
Language:
are aged <70 years old, with PSA<20ng/dL, Gleason
score ≤7
English
Publication type:
text
full-
Number of study centres:
one
Setting: hospital
Country:
Italy
Exclusion criteria:
Those with previous hormone blockade therapy or
any previous prostatic bladder neck, urethral or
pelvic surgery and total prostate volume ≥60mL ;
patients without an indwelling catheter.
Recruitment/treatment
dates: not reported
Randomisation method:
consecutive and agematched patients
randomised using computer
generated randomisation
table
Length of follow up: not
reported
time, discharge time, catheterisation,
blood loss, mobilisation, oral feeding)
Efficacy
margins, pathological stage
Nerve sparing:
Unilateral: 11/60 (18.3%)
Bilateral: 25/60 (41.7%)
A
Prospective/retrospective
data collection:
prospective
was performed with 8-10, 3-0 interrupted
sutures performed intra corporeally after
insertion of a metal bougie to expose the
urethral stump; transperitoneal route was
used.
B
Pelvic lymphadenectomy: 24/60 (40%)
120
B. Open prostatectomy: performed by
anatomic technique ; a xenon head light
and 2.5 magnification loops were used.
The urethra-vesicle anastomosis was
performed with 8-10, 3-0 interrupted
sutures with a 5/8 needle.
Quality of life
pain
Patient enrolled, n
Patient randomised, n
120
Patient analysed, n
60
60
62.29
2.9 (7.4)
Nerve sparing:
Age, mean (SD)
Unilateral: 8/60(13.3%)
Source of funding: not
reported
(8.2)
Bilateral: 31/60(51.7%)
PSA, ng/ml, mean
(SD)
6.9
6.5
(2.9)
(3)
Pelvic lymphadenectomy: 27/60 (45%)
Clinical stage, n (%)
For both A and B
66
T1
45(75%)
50(83%)
T2
15(25%)
10(17%)
Digital rectal examination, TRUS, abdominal
computed tomography scan and bone scan used for
staging.
Author, year: Jacobsen
2007
Language: English
Inclusion criteria: all men with clinically localised
prostate cancer scheduled for radical prostatectomy
(open, retropubic or laparoscopic) at the University
of Alberta.
Lymph node dissection was performed
when total serum PSA level was ≥
10ng/mL and/or Gleason score=7.
Nerve sparing was performed whenever
possible according to pre-operative
parameters such as age, clinical stage and
pre-operative potency (recorded by the
International Index of Erectile
Function(IIEF) questionnaire and Penile
Power Doppler Ultrasound evaluation)
(data not reported)
A. Laparoscopic prostatectomy
Efficacy
Approaches: trans-peritoneal.
margins, PT stage, pathological Gleason
score
No lymph node dissection.
Publication type: full-text
Number of study centres:
1
Setting: hospital
Exclusion criteria: prior pelvic radiotherapy, a
stated subjective complaint of incontinence at
baseline or a neurological impairment known to
affect bladder function.
Recruitment/treatment
dates: October 1999-July
2002
Patient recruited
consecutively, Y/N: not
urinary incontinence
Approaches: trans-peritoneal.
Lymph node dissection was conducted
when indicated.
Country: Canada
Prospective/retrospective
data collection: prospective
Dysfunction
B. Open prostatectomy
st
A (1
half)
Patients, n
67
Lost to
10 (12%)
follow-up at 1
year
nd
A (2
half)
Quality of life
B
172
24 (13%)
Additional information: patients with
risk factor for lymphatic metastases (PSA
≥20ng/ml, clinical stage ≥T3, Gleason 810) were offered open procedure in lieu of
laparoscopic procedure.
67
reported
Patients, n
29
28
148
Length of follow up: 12
months
Age, mean
(SD)
62.3 (6.4)
60.9 (6.6)
63.7 (5.7)
Source of funding: the
Northern Alberta Urology
Foundation & Alberta
Heritage Foundation for
Medical Research
BMI, mean
(SD)
26.87 (2.4) 27.54 (2.8) 28.1 (4.0)
PSA, mean
(SD)
6.9 (2.0)
7.2 (3.0)
9.8 (8.2)
0
0
2 (2%)
15 (56%)
16 (57%)
61 (49%)
8 (29%)
8 (29%)
41 (33%)
3 (11%)
0
8 (6%)
1 (4%)
4 (14%)
12 (10%)
0
0
1 (0.8%)
6.5 (0.51)
6.4 (0.64)
6.4 (0.77)
Clinical
stage,n (%)
T1b
T1c
T2a
T2b
T2c
T3a
Biopsy
Gleason
score, mean
(SD)
Author, year: Jurczok 2007 Inclusion criteria: clinical locally confined prostate
carcinoma that had been confirmed histologically.
Language: English
A. Laparoscopic prostatectomy: Preperitoneal technique with pelvic
lymphadenectomy
Publication type: full text
Safety
op conversion, surgical complications, op
time, hospital stay, catheterisation, blood
loss
Exclusion criteria: not reported
Number of study centres:
B. Open prostatectomy: Ascending
68
1
A
B
Setting:
Patient, n
163
240
Country: Germany
Age, median
(range)
62.9 (42-74)
64.8 (52-76)
PSA, ng/ml,
median (range)
7.9 (2.4-10.2)
7.25 (4.4-11.3)
Recruitment/treatment
dates: January 2003 to
April 2006
Prospective/retrospective
data collection: prospective
Clinical stage, n
(%)
Patient recruited
consecutively, Y/N: not
reported
T1a
Length of follow up: not
reported
T2a
0
6 (2.5%)
79 (48%)
75 (31.3%)
14 (8.6%)
12 (5%)
7 (4.3%)
7 (2.9%)
63(38.7%)
140 (58.3%)
retropubic technique as described by
Walsh with pelvic lymphadenectomy.
Efficacy
margins, PT stage, pathological Gleason
score
T1c
T2b
Source of funding: not
reported
Author, year: Martorana
2004
Not reported
Biopsy Gleason
scores, median
5.7
5.3
Prostate size, ml,
mean (range)
37 (18-72)
42.3 (20-120)
A. Laparoscopic prostatectomy:
Performed according to the Montsouris
technique
Inclusion criteria: not reported
Language: English
Safety
op conversion, surgical complications, op
time, hospital stay, catheterisation
Exclusion criteria: not reported
Publication type: full-text
B. Open prostatectomy:
A
Number of study centres:
B
Efficacy
69
1
Patient, n
50
50
Setting: hospital
Age, median, SD
64.6, 7.54
66.9, 5.46
Country: Italy
PSA, ng/ml,
median, SD
10.85, 9.02
13.62, 10.53
Recruitment/treatment
dates: March 2002 to
November 2003
Learning curve
op time
Clinical stage, n
(%)
Prospective/retrospective
data collection: not
reported
T1
Patient recruited
consecutively, Y/N: yes
T3
Length of follow up:
reported
Biopsy Gleason
5.56, 1.28
scores, median, SD
not
margins, PT stage, pathological Gleason
score
27 (54%)
20 (40%)
22 (44%)
27 (54%)
1 (2%)
3 (6%)
T2
5.68, 1.35
Source of funding: not
reported
Author, year: Namiki
2005
Inclusion criteria: Newly diagnosed prostate cancer
T1-T3N0M0
Language: English
Publication type: full-text
A. Laparoscopic prostatectomy:
Performed using Guillonneau &
Vallancien technique with minor
modifications.
Exclusion criteria: PSA failure greater than
0.1ng/mL within 12 months following surgery.
Number of study centres:4
A
B
Patient, n
45
121
Age, mean,
64.7, 64, 5.8
66.5, 67, 5.8
PT stage, pathological Gleason score
Dysfunction
B. Open prostatectomy: Performed using
the Walsh technique.
Setting: hospital
Efficacy
urinary function, sexual function
Country: Japan
Recruitment/treatment
dates: January 2002 to
Unilateral
A n(%)
B n(%)
21(47)
71(59)
Quality of life
70
April 2003
median, SD (range) (54-75)
Prospective/retrospective
data collection: Prospective
Co-morbidities
Patient recruited
consecutively, Y/N: not
reported
Length of follow up: not
reported
Source of funding: not
reported
(49-78)
Bilateral
3 (6)
Non-nerve sparing 21(47)
Diabetes
5
7
Cardiovascular
3
9
Other Cancer
4
10
Hypertension
9
33
Gastrointestinal
5
23
PSA, ng/ml, mean, 8.3, 7.3, 4.5
median, SD (range) (2.3-26)
20 (16)
30 (25)
Indications for nerve sparing depended on
preoperative and intraoperative factors,
prioritising cancer control
8.9, 7.3, 5.8 (254)
Clinical stage, n
(%)
T1
27 (60%)
61 (50.4%)
18 (40%)
55 (45.5%)
0
5 (4.1%)
T2
T3
Biopsy Gleason
scores, n
≤6
19 (42.2%)
48 (39.7%)
26 (57.8%)
73 (60.3%)
7
Author, year: Namiki 2006
Language: English
Inclusion criteria: patients with localised prostate
cancer.
A. Laparoscopic prostatectomy: N=64
Efficacy
pathological Gleason score
71
Publication type: full-text
Number of study centres:4
Setting: hospital
Exclusion criteria: only patients with preoperative
health related quality of life (HRQOL) data and data
from a least 2 later time points were included in the
analysis.
B. Open prostatectomy:
B1: Retropubic n=218
Dysfunction
B2: Perineal n=65
urinary function, sexual function
Nerve sparing, n (%)
Quality of life
Country: Japan
Recruitment/treatment
dates: April 2003 to March
2004
Prospective/retrospective
data collection: prospective
Patient recruited
consecutively, Y/N: not
reported
Length of follow up: 1 year
Source of funding: study
supported by a grant from
the Suzuki Urological
Foundation and the
Japanese Ministry of Health
& Welfare.
A
B1
B2
Patient, n
64
218
65
Age, mean,
median, SD
(range)
64.7, 64, 67.1, 67,
5.8
5.6
68.6, 70,
5.5
(54-77)
(56-78)
Unilateral
28 (44%) 105 (37%)
7.9, 6.8
4.4 (2.525.4)
Bilateral
3 (5%)
PSA, ng/ml,
mean, median,
SD (range)
(49-78)
10.1, 8.9, 11.8, 8.4,
6.3 (2.3- 10.6 (2.832)
67)
A
B
39 (1%)
Non-nerve sparing 33 (51%) 139 (49%)
Indications for nerve sparing depended on
preoperative and intraoperative factors,
prioritising cancer control
Clinical stage, n
(%)
T1
33 (51%) 97 (44%)
46 (71)
28 (44%) 91 (42%)
18 (28%)
3 (5%)
1 (1%)
T2
T3
30 (14%)
Biopsy Gleason
scores, n (%)
≤6
20 (31%) 47 (22%)
18 (28%)
44 (69%) 171 (78%)
47 (72%)
7
72
Author, year: Poulakis
2007
Language:
Inclusion criteria: patient who underwent extra
A. Laparoscopic prostatectomy:
peritoneal laparoscopy and pelvic lymphadenectomy
since Jan 2004 for clinically localized prostate cancer Group 1(aged ≥ 71years old) n= 72
English
Publication type: full-text
Exclusion criteria: patient with follow up of < 6
months
Number of study centres:
one
A
B
Group 1
Group 2
Patients, n
72
132
70
Age,
mean(SD)
74.1
(2.3)
57.3(2.2)
74(1.9)
BMI,
mean(SD)
29(4)
27(5)
30(5)
Previous
abdominal or
pelvic
surgery, n
18
41
(25%)
(31%)
PSA,
mean(SD)
13.5
9.1 (7.1)
Setting: hospital
Country: Germany
Recruitment/treatment
dates: 01/2004 and
07/2000
Prospective/retrospective
data collection:
retrospective
Patient recruited
consecutively, Y/N: not
reported
Length of follow up:
Source of funding: not
reported
Group 2(aged ≤ 59 years old) n= 132
surgical complications, op time, hospital
stay, catherisation, blood loss,
mobilisation, oral feeding
Nerve sparing, n (%)
Efficacy
Unilateral:
margins, PT stage, Path Gleason, PSA
recurrence
Group 1= 13(18%); Group 2= 41(31%)
Bilateral:
Group 1= 2(2.8%)
17
(24.3%)
13.7
(6.8)
Safety
Dysfunction
urinary incontinence
Group 2= 30(22.7%)
Death (none)
B. Open prostatectomy:
(6.4)
(historical cohort from 07/2000) n= 70
Clinical
stage, n (%)*
Nerve sparing, n (%)
Total
51
133
53
T1c
6(12%)
33(25%)
6 11(%)
T2a/b
27(53%)
64(48%)
30(57%)
Unilateral= 11(5.7%)
Bilateral= 3(4.3%)
73
T2c
18 35%)
36(27%)
17 32%)
Biopsy
Gleason
score,
median
(range)
7(5-9)
6(5-9)
7(5-9)
Prostate size,
cm3
mean(SD)
51(14)
47(16)
53(15)
Comorbidity,
mean(range)
2(1-2)
1(1-3)
2(1-2)
Only Group 1 was compared to
comparable cohort who underwent open
prostatectomy
*Data missing not reported in the study
Author, year: Remzi 2005 Inclusion criteria: Histologically confirmed
adenocarcinoma of the prostate and clinically ≤T2.
Language: English
Publication type: full-text
Exclusion criteria:
Number of study centres:
1
Setting: not reported
Demographic characteristics
Country: Austria
Recruitment/treatment
dates: January 2002 to
October 2003
Prospective/retrospective
data collection: prospective
Patient recruited
A1
A2
B
Patients, n
39
41
41
Age,
mean(SD)
61(11)
59(12)
60(14)
PSA,
ng/mL,
mean (SD)
5.5(3.7)
8.1(6.1)
6.9(4.4)
A. Laparoscopic prostatectomy: n=80
Safety
Cutting and dissection performed using a
harmonic scalpel (Ultracision, Ethicon,
USA) and bipolar forceps (Ethicon, USA).
A voice-controlled robotic arm (AESOP®,
Computer Motion, USA) was used for
camera guidance.
op conversion, op time, hospital stay,
surgical complications, catheterisation,
blood loss
A1Transperitoneal approach
(n=39)performed using Guillonneau &
Vallancien technique. 37/39(95%) had
staging lymphadenectomy
A2Extraperitoneal approach
(n=41)performed using Bollens et al
technique 41/41(100%) had staging
lymphadenectomy
Efficacy
margins, PT stage, pathological Gleason
score
Dysfunction
urinary continence
Quality of life
74
consecutively, Y/N: yes
Length of follow up: at
least 12 months, mean 14.9
months
Source of funding: not
reported
Gleason
score,
mean (SD)
5.1(1.2)
Prostate
size, mL,
mean(SD)
37(16)
5.5(1.3)
4.7(1.5)
B. Open prostatectomy: n=41
29/41 (71%) had staging
lymphadenectomy.
32(14)
44(18)
Nerve Sparing, n (%)
A
B
46
(57.5%)
29 (71%)
Non nerve sparing 34
(42.5%)
12 (29%)
Nerve sparing
Author, year: Salomon
2002
post-op pain
Inclusion criteria: PSA <10ng/mL
A. Laparoscopic prostatectomy: n=155
Safety
blood transfusion, op time, hospital stay,
catheterisation, surgical complications
Language: English
Exclusion criteria: not reported
B. Open prostatectomy:n=151
Publication type: full-text
Number of study centres:
1
Setting: hospital
A
B
B1 retropubic n=86
Patient, n
155
151
B2 perineal n=65
Age, mean
63.5
Retropubic,
63.8; Perineal
65.9
Efficacy
Country: France
Recruitment/treatment
dates: 1988 to 2001
Prospective/retrospective
data collection:
retrospective
PSA, ng/ml, mean 6.6
Clinical stage, n
(%)
Retropubic,
5.5; Perineal
6.5
margins, PT stage, pathological Gleason
score, PSA recurrence
Lymphadenectomy
Retropubic =all
Perineal= preoperative Gleason score >=7
Laparoscopic= preoperative Gleason
score >=7
Patient recruited
75
consecutively, Y/N: not
reported
Length of follow up: Mean
(range) retropubic, 4.7 years
(0.27 - 13.9); perineal, 5.4
years (1.7-8.6); laparoscopic
1.3 years (0.1-3.5)
Source of funding: not
reported
Author, year: Silva 2007
Language: English
T1a-b
7 (5%)
15 (10%)
T1c
106 (69%)
71 (47%)
T2a
40 (26%)
57 (38%)
T2b
2 (1%)
8 (5%)
Biopsy Gleason
scores, Mean
5.7
Retropubic,
5.6; Perineal
5.7
Inclusion criteria: patients with PSA≤15ng/ml,
Gleason score≤7 in the prostate biopsy; patients with
maximum clinical stage of T2.
Publication type: full-text
Number of study centres:
2
A. Laparoscopic prostatectomy
Efficacy
margins, PT stage, pathological Gleason
score
B. Open prostatectomy
Exclusion criteria:
A
B
90
89
Setting: hospital/private
practice
Patient , n
Country: Brazil
Age, median (range) 63(46-78)
63(46-76)
Recruitment/treatment
dates: A:05/2000-08/2004
PSA, ng/ml, median 7.36
7.99
Detail of interventions not reported.
Variance for values not specified
B: 06/1999-10/2003
Prospective/retrospective
data collection:
retrospective
Patient recruited
consecutively, Y/N: Y
76
Length of follow up: none
Source of funding: Not
reported
Author, year: Soderdahl
2005
Inclusion criteria: patients with newly diagnosed
clinically localised prostate cancer
A. Laparoscopic prostatectomy
Efficacy
PT stage
Language: English
B. Open prostatectomy
Publication type: full-text
Exclusion criteria: not reported
Dysfunction
Number of study centres:
1
A
B
116
186
A
Setting: medical centre
Complete survey data 93
86
Age, median
61
59
PSA, ng/ml, median
5.71
6
Country: US
Recruitment/treatment
dates: 2001-2003
urinary function, sexual function
Nerve sparing, n (%)
Patient , n
B
Unilateral
16 (17%) 23 (27%)
Bilateral
20 (22%) 38 (44%)
Non-nerve sparing 57 (61%) 25 (29%)
Prospective/retrospective
data collection:
prospective
Clinical stage, %
T1c
81.7%
84.9%
Patient recruited
consecutively, Y/N: not
reported
T2
18.3%
15.1%
Length of follow up: 12
months
≤6
74 (79.6%)
58 (67.4%)
7
16 (17.2%)
22 (25.6%)
8-10
3 (3.2%)
6 (7.0%)
Lost to follow up
23
100
Source of funding: US
Army & the Department of
Defence
Gleason score, n (%)
77
Analysed, n
Author, year: Terakawa
2008
Language: English
93
86
Inclusion criteria: patients who underwent both
systematic TRUS-guided needle biopsy of the
prostate and radical prostatectomy without any
neoadjuvant therapies
A. Laparoscopic prostatectomy: n=137
Efficacy
Nerve sparing:
margins, PT stage
Unilateral: 13 (9.5%)
Publication type: full-text
Bilateral: 17(12.4%)
Number of study centres:
one
Exclusion criteria:
The surgical procedure described
elsewhere
A
B
Patient, n
137
220
Age, mean (SD)
67.3 (5.8)
69.1(5.9)
B. Open prostatectomy: n= 220
PSA, ng/ml,
mean(SD)
10.9 (8.5)
12.9 (15.1)
Nerve sparing:
Setting: hospital
Country: Japan
Recruitment/treatment
dates: January 2000-April
2007
Unilateral: 19 (8.6%)
Prospective/retrospective
data collection:
retrospective
Clinical stage, n (%)
T1c
51(37%)
74(34%)
Patient recruited
consecutively, Y/N: not
reported
T2
86(63%)
146(66%)
Length of follow up: none
Biopsy Gleason
score, mean (SD)
6.5(0.9)
6.4(1.3)
Bilateral: 17 (7.7%)
The surgical procedure described
elsewhere.
Source of funding: not
reported
Digital rectal examination, transrectal
ultrasonography, PSA assay, TRUS-guided needle
biopsy, pelvic CT and bone scan was used for
78
staging.
Author, year: Touijer
2007
Language:
Inclusion criteria: Men with clinically localized
(cT1-cT3a) adenocarcinoma of the prostate.
English
Publication type: full-text
A. Laparoscopic prostatectomy: n=485
Efficacy
Performed using modified Montsouris
technique
margins, PT stage, pathological Gleason
score
Exclusion criteria: those receiving hormone therapy
prior to surgery (n=36/1213 excluded)
Number of study centres:
one
Nerve sparing:
Unilateral preservation =6%
Setting: hospital
Baseline characteristics:
Bilateral preservation=89%
Country: USA
A
B
Bilateral resection=5%
Recruitment/treatment
dates: January 2003-June
2005
Patient enrolled, n
Prospective/retrospective
data collection: prospective
Patient analysed, n
1213
B. Open prostatectomy: n=692
Standard technique
485
Patient recruited
consecutively, Y/N: Y
Age, median (IQR)
Nerve sparing:
60(55-65)
59(54-64)
5.3
5.3
(4.0-7.5)
(4.1-7.1)
T2
348(71.7%)
451(65%)
T3
125(25.8%)
213(31%)
Length of follow up: none
Source of funding:
National Cancer Institute
692
PSA, ng/ml,
median(IQR)
Unilateral preservation=6%
Bilateral preservation=91%
Bilateral resection=3%
Clinical stage, n
T1c
79
12(2.5%)
28(4%)
307(63%)
405(59%)
151(31%)
228(33%)
27(6%)
57(8%)
Biopsy Gleason
score, frequency
6
7
8-9
Partin probability of 0.37
non-organ confined
disease. Median(IQR) (0.33-0.51
0.45
(0.33-0.62)
MRI was used for clinical staging.
Author, year: Wagner
2007
Inclusion criteria: patients undergoing
prostatectomy
A. Laparoscopic prostatectomy
Safety
Montsouris technique was used.
op time, surgical complications, blood
loss
Language: English
Publication type: full-text
Exclusion criteria: not reported
B. Open prostatectomy
Number of study centres:1
A
Setting: Institution
n
75
75
Country: US
Age, mean (SD)
58 (6.9)
59 (6.9)
Recruitment/treatment
dates: not reported
BMI, mean (SD)
27 (3.0)
29 (4.5)
PSA, mean (SD)
6.2 (4.22)
8.1 (6.27)
Prospective/retrospective
data collection:
prospective
Clinical stage, n (%)
T1c
47 (63%)
Efficacy
B
45 (60%)
Anatomic approach of Walsh et al. was
used.
Nerve sparing, n (%)
A
margins, PT stage
Dysfunction
B
Unilateral
22 (29%) 9 (12%)
Bilateral
47 (63%) 62 (83%)
urinary incontinence, erectile dysfunction
Patient recruited
80
consecutively, Y/N: not
reported
Length of follow up: mean,
in total, >2 years; A, 26
months, B, 27 months
Source of funding: not
reported
T2a
21 (28%)
24 (32%)
T2b-2c
7 (9%)
6 (8%)
T3
0
0
61 (81%)
48 (64%)
12 (16%)
23 (31%)
2 (3%)
4 (5%)
Biopsy Gleason
score, n (%)
≤6
7
8-10
*author admitted there was a selection bias.
81