Prognostic Indicators of Delayed Surgical Recovery in Patients
Undergoing Cardiac Surgery
Thalita Gomes do Carmo, RN, MS, PhD1
, Rosimere Ferreira Santana, RN, MS, PhD2,
Marcos Venicios de Oliveira Lopes, RN, MS, PhD3
, Marília Mendes Nunes, RN, MS4
4
Camila Maciel Diniz, RN, MS
, Eneida Rejane Rabelo-Silva, RN, MSc, ScD5
,&
Ana Carla Dantas Cavalcanti, RN, MS, PhD2
,
1 Adjunct Professor, Undergraduate and Graduate Nursing Program, Federal Fluminense University, Rio de Janeiro, Brazil
2 Associate Professor, Undergraduate and Graduate Nursing Programs, Federal Fluminense University, CNPq Researcher, Rio de Janeiro, Brazil
3 Associate Professor, Undergraduate and Graduate Nursing Programs, Federal Ceara University, CNPq Researcher, Fortaleza, Ceara Brazil
4 PhD student, Post-Graduate Program in Nursing at Federal Ceara University, Fortaleza, Ceara Brazil
5 Associate Professor, Undergraduate and Graduate Nursing Programs - CNPq Researcher - Hospital de Clínicas de Porto Alegre - Cardiology Division,
Vascular Access Program, Universidade Federal do Rio Grande do Sul CNPq, Porto Alegre, Rio Grande do Sul Brazil
Key words
Perioperative nursing, standardized nursing
terminology, survival rates, thoracic surgery
Correspondence
Prof. Thalita Gomes do Carmo, Rua Dr.
Celestino, 74, 6º andar, Niterói, Rio de Janeiro,
Brazil.
E-mail: thalitado@gmail.com
Accepted February 13, 2021
doi:10.1111/jnu.12662
Abstract
Purpose: The purpose of this study was to analyze the prognostic capacity
of the clinical indicators of a delayed surgical recovery nursing diagnosis
throughout the hospital stay of patients having cardiac surgery.
Design: A prospective cohort design was adopted. A sample of inpatients
undergoing elective cardiac surgery was followed during the immediate
preoperative period and hospitalization. This research was conducted in
the southeast region of Brazil at a national reference institution that treats
highly complex diseases and performs cardiac surgeries. Data were collected
from July 2017 to July 2018.
Methods: At the end of 1 year of data collection, 181 patients were followed in this study. The Kaplan-Meier method was used to calculate the
survival time related to delayed surgical recovery. In addition, an extended
Cox model of time-dependent covariates was adjusted to identify the clinical signs that influenced the change in the nursing diagnosis status.
Results: A delayed surgical recovery nursing diagnosis was present in 23.2%
of the sample studied. With an expected length of stay of 8 to 10 days,
most new cases of delayed surgical recovery were observed on the 10th
postoperative day, and the survival rate after this day was decreased until
the 29th postoperative day, when the nursing diagnosis no longer appeared.
Interrupted healing of the surgical area, loss of appetite, and atrial flutter
were indicators related to an increased risk for delayed surgical recovery.
Conclusions: Timely recognition of selected clinical indicators demonstrates
a promising prognostic capacity for delayed surgical recovery.
Clinical Relevance: Accurate identification of prognostic factors allows
nurses to identify early signs of postoperative complications. Consequently,
the professional can develop an individualized plan of care, aiming at the
satisfactory clinical recovery of the patient.
Complications during the postoperative period are commonly described among patients undergoing cardiac
surgery, for example, complications such as delirium
(18.4%; Li et al., 2015), atrial fibrillation (AF; 5% to
40%) (Montrief, Koyfman, & Long, 2018), pulmonary
complications (32%), renal failure (10%), neurological
complications (7%), acute respiratory failure (3.5%;
Eremenko & Zyulyaeva, 2019), and multiorgan failure
(2%; Ariyaratnam et al., 2019).
428
These complications can lead to the following sequelae for patients: increased hospital length of stay; delayed
social activities; decreased quality of life; traumatic
memories related to the frustrated expectation of recovery; and deleterious effects for the institution, such
as
increased
hospital
costs
(Almashrafi
&
Vanderbloemen, 2016; Engelman et al., 2019; Pimentel
et al., 2017; Quintana & Kalil, 2012). The nursing
care plan requires precise diagnostic inference and,
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Gomes do carmo et al.
consequently, early interventions to minimize the possible damage related to prolonged hospitalization
(Engelman et al., 2019).
In this context, the nursing diagnosis of delayed
surgical recovery is defined in the NANDA International
(NANDA-I) taxonomy as an "extension of the number
of postoperative days required to initiate and perform
activities that maintain life, health, and well-being"
(Herdman & Kamitsuru, 2018, p. 803).
Prognostic factors are useful to understand how specific clinical markers affect the probability of a patient
having a nursing diagnosis of interest. An accurate
identification of prognostic factors allows nurses to
identify early signs of postoperative complications.
Consequently, the professional is able to develop an
individualized nursing care plan, aiming at a satisfactory or optimal patient outcome.
Diagnostic survival information obtained from prognostic studies can be used to facilitate the diagnostic
inference made by nurses, since it points out which
clinical indicators are good predictors of a specific nursing diagnosis (Lopes, Silva, & Araujo, 2012). Studies
about prognostic factors allow for the examination of
the effectiveness of nursing interventions, since a followup is required (Lopes et al., 2012). It is also possible
to identify the time of diagnostic survival in these
studies, which is defined as the time to occurrence
of an event or outcome of interest, which, in this
study, is the nursing diagnosis delayed surgical recovery.
The survival time is defined as the amount of time
that an individual remains in a well-defined state
(without delayed surgical recovery), from an event
(surgery), until a second event is also observed (manifestation of delayed surgical recovery; Fletcher, Fletchert,
& Fletchert, 2013).
The study of prognostic factors helps to identify the
degree of prediction of these factors by assessing the
manifestations presented by the patient. Therefore,
nurses can predict the evolution of an outcome based
on the clinical course or the natural history of the
disease (Elmore, Wild, Nelson, & Katz, 2020).
In nursing, prognostic studies aim to evaluate nurses’
clinical judgment when they are faced with the evolution of a nursing diagnosis based on the characteristics
and clinical conditions presented by patients and their
therapeutic responses (Lopes et al., 2012). Thus, it is
possible to identify the clinical indicators presented by
individuals who could influence the occurrence of a
nursing diagnosis (Lopes et al., 2012).
The clinical indicators of a nursing diagnosis are
understood and referred to as prognostic factors; that
is, these characteristics refer to the clinical conditions
(signs and symptoms or clinical indicators) presented
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Prognostic Indicators of Delayed Surgical Recovery
by patients that may lead to an (unintended) outcome
of the presence of the nursing diagnosis (Elmore et
al., 2020; Lopes et al., 2012).
Consequently, the following question arises: Are there
clinical indicators that can predict the delayed surgical
recovery nursing diagnosis in patients having cardiac
surgery? In view of the above, the present study aim
is to analyze the prognostic capacity of the clinical
indicators of a delayed surgical recovery nursing diagnosis among patients having cardiac surgery.
Methods
Design
This prognostic study is a clinical validation of the
delayed surgical recovery nursing diagnosis based on
the prospective cohort study methodology. The Equator
network guideline used was the strengthening the
reporting of observational studies in epidemiology
(STROBE).
Setting
This research was conducted in the southeastern
region of Brazil at a national reference institution that
treats highly complex diseases and performs cardiac
surgeries. This hospital has 15 intensive care unit (ICU)
beds that are available for postoperative care. In Brazil,
the expected length of stay following traditional cardiac
surgeries ranges from 8 to 10 days, with a mean of
3 to 4 days in the ICU and 5 to 6 days in the regular
wards (Silva, Sousa, Soares, Colósimo, & Piotto, 2013;
Silveira, Santos, Moraes, & Souza, 2016).
Participants
A sample of inpatients undergoing elective cardiac
surgery was followed during the immediate preoperative period and hospitalization. The population consisted
of patients admitted to the surgical ward who were
awaiting an elective cardiac surgical procedure. The
following inclusion criteria were considered: age greater
than or equal to 18 years; hemodynamic stability;
surgical myocardial revascularization (MRV) with or
without extracorporeal circulation, valvuloplasty, or
combined surgeries (MRV and valvular changes or
valvuloplasty); and correction of an aortic artery aneurysm. The following exclusion criteria were adopted:
a previous history of dementia or inability to understand and communicate; cardiac transplantation; or
hospitalization for nonsurgical procedures for the treatment of endocarditis.
429
Prognostic Indicators of Delayed Surgical Recovery
The sample size was determined assuming the relevant delayed surgical recovery nursing diagnostic
survival rates, since these values are unknown in
practice. Thus, we assumed a survival rate (S1 [∞])
of 50% of patients with a certain clinical indicator
or defining characteristic, and a difference of 20%
for the survival between patients without a clinical
indicator or defining characteristic to obtain a value
for S2 (∞) of 70%. In addition, a confidence level
of 95% and a statistical power of 80% were specified and applied to the formula recommended by
Lee and Wang (2013) for studies identifying prognostic factors. With these values, the number of
participants needed for the study was 181.
Data Collection Procedures
This study has been approved by the Research Ethics
Committee at the Antonio Pedro University Hospital
and at the National Cardiology Institution. Initially, we
reviewed the surgical schedule to select potentially
eligible patients. Once potential patients were identified,
each patient was approached at bedside by one of the
researchers from the research team who explained the
study and its purpose to the patient, read out loud
the consent form, and answered any questions.
After the surgery, each patient was observed by one
of the members of the research team, composed of
six baccalaureate nursing students nearing their program completion and one doctoral nursing student.
Both the baccalaureate students and the doctoral student were members of a research group and familiar
with the NANDA-I taxonomy. Observations were made
throughout the hospitalization until discharge to another
institution or death. The research team reviewed each
patient’s medical record daily, and the data collection
was carried out on all 7 days of the week until the
sample size was reached.
The assessment of the defining characteristics and
related factors of delayed surgical recovery was made
by a research team composed by undergraduate nursing students and a PhD student, who are members
of a research group on nursing taxonomies. The research
team was trained prior to data collection to assess the
patients and collect the data in a standardized manner. This training included the discussion of clinical
cases and statistical simulations of false-positive, falsenegative, efficiency, and tendency situations, as recommended by Lopes et al. (2012).
Variables
The clinical indicators listed by the NANDA-I taxonomy for delayed surgical recovery include discomfort,
evidence of the interrupted healing of the surgical
area (EIHSA), excessive time required for recuperation,
impaired mobility, inability to resume employment, loss
430
Gomes do carmo et al.
of appetite, postponement of the resumption of work,
and need for assistance with self-care (Herdman &
Kamitsuru, 2018). Conceptual and operational definitions were used as part of the data collection protocol,
with the purpose to better define and measure the
constructs, and to enhance the consistency of the data
collection (Carmo, Santana, Antunes, & Carvalho, 2018).
These definitions were developed based on the NANDA-I
taxonomy and from a literature review of institutional
protocols available in articles, books, and manuals as
described below:
1. Discomfort: Determined from the physical examination and verbal report of the patient using a scale
from 0 (absence of discomfort) to 10 (extreme discomfort).
For the analysis, this variable was dichotomized as
0 (absent) or 1 (present; Mendonça & Andrade, 2015).
2. Pleural effusion: Determined from the medical records,
being classified as present only when the effusion
was severe and required drainage (Feller-Kopman
et al., 2018).
3. Malignant arrhythmia: Determined from the medical
records, being classified as present when resulting
in cardiorespiratory arrest (ventricular fibrillation and
asystole; Jason, Timothy, Richard, & Spencer, 2017).
4. Evidence of interrupted healing at the surgical area:
Determined from physical examination or the medical
records. This clinical indicator was considered present
when a delay in the process of healing could be
observed (e.g., red surgical wound, indurated surgical
wound, and presence of signs of draining; Appoloni,
Herdman, Napoleão, Carvalho, & Hortense, 2013;
Gallafrio et al., 2017).
5. Surgical site infection: Determined by the Magedanz
SCORE tool and when patients had at least two or
more of the following signs and symptoms: fever
(body temperature >38°C), chest pain or sternal
instability, and purulent draining at the surgical areas
(Oliveira et al., 2017).
6. Atrioventricular block (second degree or higher):
Determined from the medical records (Souza &
Scanavacca, 2016).
7. Loss of appetite: Determined from the Mini Nutritional
Assessment (MNA) test. The full MNA is a validated
screening tool that identifies persons who are malnourished or at risk for malnutrition. The maximum
score is 14, and a score of 12 to 14 points is classified as normal nutritional status, while a score of
8 to 11 points is classified as risk for malnutrition,
and a score of 0 to 7 points is classified as malnutrition. This variable was assessed every 3 days during
the postoperative period until discharge (Bolado,
Fernádez, Muñoz, & Aller, 2019).
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Gomes do carmo et al.
8. Need for assistance with self-care: Determined from
the Barthel Index. The Barthel Index uses 10 variables describing activity of daily living and mobility,
and it is related to requiring help to complete activities such as feeding, bathing, dressing, transferring,
grooming, and others (Minosso, Amendola, Alvarenga,
& de Campos Oliveira, 2010). Shah, Vanclay, and
Cooper (1989) suggested that scores from 0 to 20
indicate total dependency, scores from 21 to 60 indicate severe dependency, scores from 61 to 90 indicate
moderate dependency, and scores from 91 to 99
indicate slight dependency. We considered the variable requires help to complete self-care as present
when the patient scored 21 or greater on the Barthel
Index and absent when the score was equal to or
less than 20.
9. Benign arrhythmias: Determined from the medical
records, also called atrial flutter, which, although
having a transient and self-resolving character, is
associated with complications such as increased risk
for stroke and mortality, prolonged hospital stay, and
increased costs (Rosso & Viskin, 2020; Souza &
Scanavacca, 2016).
10.
Impaired mobility: Determined from the Barthel
Index component related to mobility. This indicator
was measured from the Simplified Barthel Index (SBI)
using the following cutoff values: Category 1 = high
dependency: SBI score = 0–11; Category 2 = mild
dependency: SBI score = 12–17; and Category 3 =
low dependency: SBI score = 18–20) (Lam, Lee, &
Yu, 2014; Minosso et al., 2010). The indicator was
considered present in patients classified into categories
1 or 2.
The clinical indicators inability to resume employment and postponement of the resumption of work
were not assessed. These indicators occur postdischarge
and were outside the parameters of this study.
Data Analysis
The data were analyzed using the Statistical Package
for the Social Science program (SPSS) version 22.0
(IBM Corp., Armonk, NY, USA) and R software version 3.2.1 (R Foundation for Statistical Computing,
Vienna, Austria). The descriptive statistics included
proportions, minima, maxima, means, medians, standard
deviations, and coefficients of variation. The
Kolmogorov-Smirnov test was used to verify the adherence of the data to a normal distribution.
For the determination of the diagnostic decisions of
delayed surgical recovery, a latent class analysis was
used with the application of random effects for all
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Prognostic Indicators of Delayed Surgical Recovery
the observations of the sample. Studies that analyze
the accuracy of nursing diagnoses suffer from a lack
of a gold standard, since some clinical indicators presented by individuals cannot be measured directly by
devices. Several clinical indicators are subjective and
correspond to particular human responses that cannot
be standardized (Lopes et al., 2012). Besides that, many
nursing diagnoses from the NANDA-I taxonomy are
general and there is no set of defining characteristics
to specific groups such as newborns, elderly, and children, for example.
The latent class analysis is a robust and useful
method for conducting nursing diagnostic accuracy
studies, in the absence of a gold standard and the
particularities described above. This type of analysis
allows researchers to create a model that indicates
which clinical indicators, combined, are related to
the presence of the nursing diagnosis of interest,
thus being considered the best (most accurate)
predictors.
To overcome this limitation, a latent class analysis
was used to obtain accuracy measurements. This analysis
is used to calculate measurements of accuracy when
there is no perfect reference pattern and is based on
the supposition that an unobserved or latent variable
(nursing diagnosis) determines the associations between
the observable variables (clinical indicators). The model
includes clinical indicators that presented at least one
of the diagnostic accuracy measures (sensitivity and
specificity) with values greater than 0.5 and 95% confidence intervals (CIs) that did not include this value.
The adjustment of the model was verified by applying
the likelihood ratio test (G2), in which nonsignificant
values indicate a good adjustment.
In addition to this test, the relative entropy value
of the model was calculated. Entropy indicates the
ability of a model to clearly identify individuals with
or without a diagnosis. This measurement varies between
0 and 1, and values greater than 0.8 indicate good
entropy (Lee & Wang, 2003; Pepe, 2003; Rutjes, Reitsma,
Coomarasamy, Khan, & Bossuyt, 2007; Zhou,
Obuchowsk, & McClish, 2011).
After the identification of the latent class model,
the posterior probabilities were calculated and used to
establish the presence or absence of the nursing diagnosis based on the possible clinical indicator combinations found in the sample. The nursing diagnosis was
considered present when the posterior probability was
higher than 0.5; otherwise, the nursing diagnosis was
considered absent.
The inferential analysis was based on the determination of the CIs of the estimated statistics and
tests of statistical significance. All significance tests
431
Prognostic Indicators of Delayed Surgical Recovery
were performed considering a maximum significance
level of 5%, and the following decision rule was
adopted for the tests: the null hypothesis is rejected
when the p value associated with the test is less
than .05.
The measures used for epidemiological analysis were:
number of postoperative cardiac surgery patients at
risk for developing the nursing diagnosis, number of
new cases for the nursing diagnosis from the number
of postoperative cardiac surgery patients in risk, risk
rate (calculated by the ratio between new cases and
patients at risk), person/time (refers to the sum of
time elapsed for each patient after cardiac surgery until
the development of the nursing diagnosis), and rate
of incidence (given by the ratio between the number
of new cases and the value of the person/time). The
risk rate refers to the proportion of postoperative cardiac surgery patients without a nursing diagnosis who,
on average, will develop it during the follow-up period.
The incidence rate reflects the speed with which new
cases of the nursing diagnosis studied are
developed.
The Kaplan-Meier method was used to calculate the
survival time related to delayed surgical recovery and
determine the demographic and clinical variables that
were related to survival time (Bewick, Cheek, & Ball,
2004). In addition, an extended Cox model adjusted
for the time-dependent covariates was performed to
identify the variables that influenced the change in
the nursing diagnosis’ status by changing their status
(absent or present; Cox, 1972). This type of modeling
is used when the value (quantitative variables) or
occurrence (qualitative variables) of a phenomenon
changes over time and when such a change can cause
a change in survival time.
In this model, for each variable, the risk ratios and
their respective CIs were presented. Regarding the risk
ratios, values above 1 indicate harm, and values below
1 indicate protection. In addition, a Wald test was
applied to identify the variables that influence the
survival time of the patients. The coefficients of the
model are presented to enable the calculation of the
prognostic index of delayed surgical recovery. The
adjustment of the model was analyzed by the coefficient of determination (R2), the likelihood ratio (G2)
test, the Wald test, the log-rank test, and the concordance index.
The proportional hazards assumption was evaluated
by the calculation of the Schoenfeld residuals and correlation coefficients. In this analysis, the proportional
hazards assumption is verified by the nonstatistical
significance of the correlation coefficients (rho) assessed
by the chi-square test. These analyses refer to the
432
Gomes do carmo et al.
verification of the assumption that the risks are proportional, that is, the ratio between the risk for occurrence of the event between two individuals is constant
over time and proportional to the number of prognostic
factors of each individual. Thus, statistically significant
correlations between the Martingale residuals model
or standardized residues outside the range include at
least 95% of the residuals between -2 and +2 mean
self-correlation and disproportionate risks. In addition,
graphs of the Schoenfeld and Martingale residuals are
presented.
Results
In total, 181 subjects were followed throughout their
hospital stay according to the flowchart (Figure S1).
From the 181 patients initially enrolled in the study
who did not have this nursing diagnosis preoperatively,
42 cases of delayed surgical recovery emerged during
the first 29 postoperative days.
Demographic and Baseline Clinical Characteristics
Table S1 shows the demographic and baseline clinical characteristics of the study patients. The most frequently performed cardiac surgery was myocardial
revascularization (n = 99, 54.7%), and the most frequent postoperative complication was invasive mechanical ventilation for ≥8 hr (n = 111, 61.3%).
Accuracy of the Clinical Indicators
The accuracy measurements of the clinical indicators
of delayed surgical recovery are shown in Table S2.
The main clinical indicators of delayed surgical recovery
were discomfort and evidence of the interruption of
surgical healing, with high sensitivity (1.00 and 0.88,
respectively) and high specificity (0.82 and 0.99, respectively) values. The other clinical indicators presented
high specificity values, including surgical site infection
(SSI; 0.99), pleural effusion (0.87), malignant arrhythmia (0.99), and atrioventricular block (0.99). The clinical
indicators loss of appetite and need for assistance for
self-care are not in Table S2 because they did not
have a good accuracy level.
Posterior Probabilities
In Table S3, we show the posterior probabilities,
which are sets of clinical indicators formed by the
latent class model indicating the presence or absence
of delayed surgical recovery.
According to Table S3, sets representing 16 combinations were formed; from these sets, only 6 indicated
the presence of delayed surgical recovery with posterior
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Gomes do carmo et al.
probabilities higher than the value determined by the
test (0.83) for both the presence and absence of the
phenomenon. The column "n" represents the number
of times in which patients presented the assigned set
in each row. The nursing diagnosis was present in
23.2% of the sample studied.
Incidence Rate of Delayed Surgical Recovery
Regarding the incidence rate of delayed surgical
recovery, initially, 181 patients did not have this nursing diagnosis. During the follow-up, 42 new cases
emerged. The daily incidence rate of the nursing diagnosis was 9.81% among the patients.
Survival Rates
Table S4 shows the probabilities of patients having
cardiac surgery manifesting delayed surgical recovery
over time. It is observed that on the 10th postoperative day, this probability was 83.6%, representing the
period during which the greatest number of events
occurred. Additionally, the time for a patient to develop
delayed surgical recovery was 29 days, since patients
were either discharged without having developed this
nursing diagnosis after this period or died. Thus, the
survival rate for the probability of the nursing diagnosis
delayed surgical recovery among patients undergoing
selected cardiac surgeries up to the 29th postoperative
day was 61%. This means that a reasonable number
of patients did not have delayed surgical recovery during the postoperative period, while close to 40% presented the nursing diagnosis.
Although the manifestations of delayed surgical recovery occurred until the 29th postoperative day, events
after the 10th day tended to decrease until none were
detected after the 29th day.
The Cox Model
Table S5 shows that, in the Cox model, the clinical
indicator of evidence of the interruption of surgical
healing achieved a high-risk ratio (relative risk [RR]
= 34.04, 95% CI = 17.91–64.70, p < .001) for the
nursing diagnosis delayed surgical recovery, which
means that the presence of this defining characteristic
increases the risk for a patient developing the nursing
diagnosis by more than 34 times. Other clinical indicators that were significantly related with the occurrence of the nursing diagnosis of interest were presence
of benign arrhythmias (RR = 2.81, 95% CI = 1.35–
5.87, p = .006), loss of appetite (RR = 2.62, 95% CI
= 1.44–4.79, p = .002), and need for assistance with
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Prognostic Indicators of Delayed Surgical Recovery
self-care (RR = 1.54 equals to 54% risk, 95% CI =
0.99–2.39, p = .053). Data from other clinical indicators without statistical significance are not shown in
the table.
The extended Cox model presented a good fit for
four clinical indicators that achieved the highest risk
ratios: EIHSA, benign arrhythmias, loss of appetite,
and need for assistance with self-care. Thus, the maximum value of the prognostic index corresponds to a
person within the first 10 postoperative days and having at least one of these four clinical indicators.
Discussion
After statistical analysis of all clinical indicators collected, the combination of six clinical variables was
shown to be sensitive and specific for the determination of the nursing diagnosis; these variables included
discomfort, evidence of interrupted healing of the surgical area, SSI, pleural effusion, malignant arrhythmia,
and atrioventricular block. Although studies investigating the nursing diagnosis of delayed surgical recovery
in cardiac surgery have not been performed to date,
it is known that these clinical variables indicate the
extension of the hospital stay of patients having cardiac
surgery in the literature (Carmo et al., 2018), thus
corroborating the findings of the study.
The clinical indicator discomfort was considered one
of the best predictors of delayed surgical recovery,
since it achieved high sensitivity and high specificity
(sensitivity: 1.00; specificity: 0.82). This observation is
corroborated by research concerning postoperative
patients following cardiac surgery, showing that discomfort is caused by several factors, such as frequent
monitoring of vital signs and capillary glycemia, which
cause anxiety and distress in patients (Mendonça &
Andrade, 2015; Smith & Custard, 2014).
Additionally, orotracheal intubation and mechanical
ventilation are standard practice for patients having
cardiac surgery until the effects of anesthesia have
dissipated and patients are able to breathe satisfactorily
on their own. However, while considered necessary,
Smith and Custard (2014) reported that 46.9% of
cardiac surgical patients who have experienced an
orotracheal tube reported that the tube caused discomfort due to the inability to communicate, 33.5%
of patients reported anxiety due to being awakened
after airway extubation, 13.4% of patients experienced
fear, and 6.7% of patients experienced pain. In addition, drains and catheters, which are indispensable for
postoperative treatment, cause discomfort and insecurity.
Once these objects are removed, the patients report
relief, with 46.9% of patients reporting tranquility,
433
Prognostic Indicators of Delayed Surgical Recovery
26.8% of patients reporting freedom, and 13.4% of
patients reporting pain relief (Mendonça & Andrade,
2015; Smith & Custard, 2014).
With respect to the clinical indicator evidence of
the interruption of the healing of the surgical area
(sensitivity: 0.88; specificity: 0.99), in the studies conducted by Santana et al. (2018) and Rembold, Santana,
Lopes, and Melo (2020) this indicator also presented
a high prognostic value for delayed surgical
recovery.
Although SSI (sensitivity: 0.52; specificity: 0.99) did
not show good sensitivity, this indicator demonstrated
high specificity, which is defined as the capacity to
predict the absence of the postoperative nursing diagnosis when it was absent. This correlation is observed
in a study conducted by Cowper et al. (2007), who
observed that patients with SSI had a prolonged length
of hospital stay compared to patients without the infectious manifestation, with an average hospital stay of
45 days and 9 to 18 days, respectively. The hospitalization time of patients presenting SSI may be up to
three times longer than that of patients who do not
present SSI, according to a study that retrospectively
evaluated 280 patient charts. This clinical indicator
contributes to more complications, increased hospital
stays, high hospital costs, and the suffering of patients
and their relatives (Braz, Evangelista, Evangelista,
Garbaccio, & Oliveira, 2018).
Pleural effusion (sensitivity: 0.24; specificity: 0.87)
is among the most common pulmonary complications
during the immediate postoperative period following
cardiac surgery, affecting 18% of the 228 patients
analyzed by Soares et al. (2011), and is essentially
related to the removal of the mammary artery for
revascularization surgeries, the need for orotracheal
intubation, and the need for mechanical ventilation
for more than 48 hr. These findings are similar to
those found in this study, in which pleural effusion
showed good specificity (0.87). It means that when
this clinical indicator is absent, the delayed surgical
recovery nursing diagnosis also would be absent.
Malignant ventricular arrhythmias (sensitivity: 0.00;
specificity: 0.99), which lead to cardiorespiratory arrest
(ventricular fibrillation and asystole), have an incidence
of 2% to 13%; thus, this indicator is more common
and frequent between the third and fifth postoperative
days following cardiac surgery (Souza & Scanavacca,
2016) and results in a prolongation of hospitalization
time, greater readmission rate in intensive care units,
prolonged mechanical ventilation, need for inotropic
drugs or mechanical circulatory support, and even reintubation (Greenberg, Lancaster, Schuessler, & Melby,
2017).
434
Gomes do carmo et al.
In this study, the absence of bradyarrhythmia (atrioventricular blocks, which achieved a high specificity
of 0.99 and low sensitivity of 0.00) in the postoperative period, also showed a strong relationship with
the absence of the delayed surgical recovery nursing
diagnosis. These results can be explained by the fact
that bradyarrhythmias are common events after cardiac
surgery and may be caused by "hypothermia, incomplete lavage of the cardioplegic solution (solution that
facilitates the cessation of heartbeats during surgery),
antiarrhythmic drug in the perioperative period and
trauma of the conduction system" (Reynolds et al.,
2016; Souza & Scanavacca, 2016, p. 353).
Forty-two participants (23%) experienced delayed
surgical recovery. From these 42 participants, none
died, while among the 139 participants that did not
have this nursing diagnosis, 12 died. These 12 patients
progressed to death in the first 48 hr after surgery;
thus, there was not enough time to present the clinical manifestations of delayed surgical recovery. We
believe that there is no direct correlation between the
presence or absence of the nursing diagnosis delayed
surgical recovery and death. Postoperative deaths are
common in the high-risk surgical population. This
population consist mainly of elderly patients with
comorbidities who undergo major surgeries. The nursing diagnosis of delayed surgical recovery is prevalent
in patients having chronic conditions, and who require
continued care.
Regarding the diagnostic survival rate, it was possible to infer that the 10th postoperative day was
the most critical for the occurrence of the nursing
diagnosis delayed surgical recovery. In this study conducted in Brazil, 83.6% of patients demonstrated
clinical indicators for delayed surgical recovery by the
10th postoperative day. Therefore, it can be inferred
that patient status by the 10th postoperative day is
the most critical in the identification of delayed surgical recovery. However, this finding may vary by
study sample and country. In a systematic review by
Almashrafi and Vanderbloemen (2016), delayed recovery may appear in sooner than 10 days in locations
such as the United States, Canada, Germany, and
Norway, with a lesser mean length of stay of 2 to
10 days.
In the face of this result, it can be inferred that
the 10th postoperative day requires more awareness
from the health team and interventions to avoid the
occurrence of delayed surgical recovery. In our sample,
when the hospitalization period exceeded 10 postoperative days, the patients were already diagnosed with
delayed surgical recovery and remained at the hospital
for the treatment of complications.
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Gomes do carmo et al.
Similar to this study, Silva et al. (2013) showed
that during the postoperative period following cardiac
surgery, the patients who were discharged before the
10th postoperative day did not present complications
and so did not need to prolong their stay; these patients
finished their recovery at home with scheduled outpatient returns. Another study evaluating the length
of hospital stay in 66,587 patients having myocardial
revascularization from 2007 to 2009 in 10 European
countries described an average hospital stay of 9 to
17 days (Gaughan et al., 2012).
A daily incidence rate of delayed surgical recovery
of 9.81% was observed. In addition to presenting good
sensitivity and specificity in the diagnostic accuracy
tests described above, EIHSA, which is described by
the NANDA-I, was an excellent prognostic indicator
with the highest risk ratio (RR = 34.04; CI = 17.91–
64.70). This clinical indicator was observed and recorded
in cases of surgical dehiscence.
The loss of appetite and need for assistance with
self-care indicators are also described by the NANDA-I
as clinical indicators and were good prognostic indicators, with a risk ratio of 2.6 times (CI = 1.44-4.79)
and 54% in the chance of manifesting nursing diagnosis. Loss of appetite after cardiac surgery is a common finding during the immediate postoperative period
and becomes a more serious problem when it persists
beyond these 24 to 48 postoperative hours (Koerich,
Baggio, Erdmann, Lanzoni, & Higashi, 2013). This
symptom can sometimes be characterized as a consequence of the other symptoms, such as anxiety and
pain (Mendonça & Andrade, 2015).
Studies corroborate that after the middle period,
which is marked by the removal of drains and the
beginning of a greater mobilization in bed, it is time
to encourage the patient to perform self-care activities.
This encouragement helps patients (re)insert themselves
into a routine and satisfy their basic human needs
and probable deficits in their self-care, thus rendering
them more independent of the nursing team (Smith
& Custard, 2014).
Benign arrhythmias as are not described by the
NANDA-I as a specific clinical indicator of delayed
surgical recovery but were investigated because the
literature describes them as predictors of complications;
benign arrhythmias had a risk ratio of 2.8 times (CI
= 1.35–5.87) in the participants experiencing the nursing diagnosis. Arrhythmias during the postoperative
period following cardiac surgery are usually benign
and transient (spontaneously converting to sinus
rhythm), present with a frequency of 30% in myocardial revascularization surgeries and approximately
60% of valvular surgeries, and have a higher
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Prognostic Indicators of Delayed Surgical Recovery
prevalence during the first 5 postoperative days
(Greenberg et al., 2017; Magalhães et al., 2016; Souza
& Scanavacca, 2016). Despite their transient character,
arrhythmias can extend the length of stay in the hospital, which in turn can lead to complications (including the occurrence of the nursing diagnosis delayed
surgical recovery).
The length of stay at the ICU among patients who
develop AF increases by an average of 2 to 4 days
in comparison to those who maintain sinus rhythm.
This complication is also the main cause of hospital
readmission after discharge following cardiac surgery
(Magalhães et al., 2016). Stroke, hypotension, acute
pulmonary edema, a long ICU stay, and high additional
costs are directly associated with AF (Greenberg et
al., 2017) and increased morbidity and mortality.
These findings are relevant since they can help nurses
during the diagnostic inference process in making early
and accurate interventions. Further studies would be
worthwhile to develop a scale for early detection of
delayed surgical recovery (or risk for this nursing diagnosis) based on the risk ratios and accuracy
measurements.
Conclusions
The 10th postoperative day was shown to be the
most important day in the postoperative period for
the identification of delayed surgical recovery. With
an average length of stay of 19 days, patients hospitalized on this day demonstrate a delay in the expected
recovery process.
The clinical indicators with the best predictive values
and prognostic capacity were evidence of interrupted
healing of the surgical area, loss of appetite, and atrial
flutter. These indicators were related to an increased
risk for having delayed surgical recovery per unit of
time.
Relevance to Clinical Practice
The early detection of the clinical indicators of delayed
surgical recovery is relevant to the clinical practice of
perioperative nurses since scientific evidence has confirmed that nurses’ assessment is indispensable in the
care of patients during the recovery period after cardiac
surgery.
Acknowledgments
Research entitled “Delayed Surgical Recovery after
Cardiac Surgery: Prognostic Study” was conducted with
financial support from the Institutional Program for
435
Prognostic Indicators of Delayed Surgical Recovery
Scientific Initiation Scholarships (AGIR/PROPPi/UFF) of
the National Council for Scientific and Technological
Development (CNPq).
Clinical Resource
• NANDA International. https://nanda.org/
References
Almashrafi, A., & Vanderbloemen, L. (2016).
Quantifying the effect of complications on patient
flow, costs and surgical throughputs. BMC Medical
Informatics and Decision Making, 16(1), 136. https://
doi.org/10.1186/s12911-016-0372-6
Appoloni, A. H., Herdman, T. H., Napoleão, A. A.,
Carvalho, E. C., & Hortense, P. (2013). Concept
analysis and validation of the nursing diagnosis
Delayed surgical recovery. International Journal of
Nursing Knowledge, 24, 115–121. https://doi.
org/10.1111/j.2047-3095.2013.01241.x
Ariyaratnam, P., Ananthasayanam, A., Moore, J.,
Vijayan, A., Hong, V., & Loubani, M. (2019).
Prediction of postoperative outcomes and long-term
survival in cardiac surgical patients using the
Intensive Care National Audit & Research Centre
score. Journal of Cardiothoracic and Vascular
Anesthesia, 33(11), 3022–3027. https://doi.
org/10.1053/j.jvca.2019.05.034
Bewick, V., Cheek, L., & Ball, J. (2004). Statistics
review 12: survival analysis. Critical Care, 8(5), 389.
Bolado, C. J., Fernádez, H. O., Muñoz, M. M., &
Aller, R. F. (2019). Undernutrition measured by
the Mini Nutritional Assessment (MNA) test and
related risk factors in older adults under hospital
emergency care. Nutrition, 66, 142–146. https://doi.
org/10.1016/j.nut.2019.04.005
Braz, N. D. J., Evangelista, S. D. S., Evangelista, S.
S., Garbaccio, J. L., & Oliveira, A. C. D. (2018).
Surgical site infection in patients submitted to
cardiac surgeries: An epidemiological profile
analysis. Revista de Enfermagem do Centro-Oeste
Mineiro, 8, e1793. https://doi.org/10.19175/recom.
v8i0.1793
Carmo, T. G., Santana, R. F., Antunes, B. S., &
Carvalho, M. R. (2018). Deferred surgical
recovery in adults and the elderly in
perioperative perioperative thoracic surgery:
Systematic review. Sociedad Mexicana de Cardiologia,
26(2), 22–28.
Cowper, P. A., DeLong, E. R., Hannan, E. L.,
Muhlbaier, L. H., Lytle, B. L., Jones, R. H., &
436
Gomes do carmo et al.
Peterson, E. D. (2007). Is early too early? Effect of
shorter stays after bypass surgery. Annals of Thoracic
Surgery, 83(1), 100–107. https://doi.org/10.1016/j.
athoracsur.2006.08.001
Cox, D. R. (1972). Regression models and life—
Tables. Journal of the Royal Statistical Society: Series B,
34(2), 187–220.
Elmore, J. G., Wild, D., Nelson, H. D., & Katz, D. L.
(2020). Jekel’s epidemiology, biostatistics, preventive
medicine, and public health (5th ed.). Philadelphia,
PA: Elsevier.
Engelman, D. T., Ali, W. B., Williams, J. B., Perrault,
L. P., Reddy, V. S., Arora, R. C., … Boyle, E. M.
(2019). Guidelines for perioperative care in cardiac
surgery: Enhanced recovery after surgery society
recommendations. JAMA Surgery, 154(8), 755–766.
https://doi.org/10.1001/jamasurg.2019.1153
Eremenko, A. A., & Zyulyaeva, T. P. (2019).
Postoperative acute respiratory failure in cardiac
surgery. Khirurgiya, 8, 5–11. https://doi.
org/10.17116/hirurgia20190815
Feller-Kopman, D. J., Reddy, C. B., DeCamp, M. M.,
Diekemper, R. L., Gould, M. K., Henry, T., …
Rahman, N. M. (2018). Management of malignant
pleural effusions. An official ATS/STS/STR clinical
practice guideline. American Journal of Respiratory
and Critical Care Medicine, 198(7), 839–849. https://
doi.org/10.1164/rccm.201807-1415ST
Fletcher, R., Fletchert, S. W., & Fletchert, G. S.
(2013). Clinical epidemiology: The essentials (5th ed.).
Philadelphia, PA: Lippincott Williams & Wilkins.
Gallafrio, S. T., Menezes, T. T., Monaco Filho, F.,
Alvarez, D. M., Strabelli, T. M. V., Jatene, F. B., &
Gemperli, R. (2017). Analysis of the treatment
protocol for sternotomy wound infection after
cardiac surgery. Brazilian Journal of Plastic Surgery,
32(2), 194–201. https://doi.org/10.5935/21771235.2017RBCP0031
Gaughan, J., Kobel, C., Linhart, C., Mason, A.,
Street, A., & Ward, P. (2012). Why do patients
having coronary artery bypass grafts have different
costs or length of stay? An analysis across 10
European countries. Health Economics, 21(Suppl. 2),
77–88. https://doi.org/10.1002/hec.2842
Greenberg, J. W., Lancaster, T. S., Schuessler, R. B., &
Melby, S. J. (2017). Postoperative atrial fibrillation
following cardiac surgery: A persistent complication.
European Journal of Cardiothoracic Surgery, 52(4),
665–672. https://doi.org/10.1093/ejcts/ezx039
Herdman, T. H., & Kamitsuru, S. (2018). Nursing
diagnoses: Definitions and classification 2018–2020.
Oxford, UK: Wiley-Blackwell.
Jason, W. G., Timothy, S. L., Richard, B. S., &
Spencer, J. M. (2017). Postoperative atrial
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Gomes do carmo et al.
fibrillation following cardiac surgery: A persistent
complication. European Journal of Cardio-Thoracic
Surgery, 52(4), 665–672. https://doi.org/10.1093/
ejcts/ezx039
Koerich, C., Baggio, M. A., Erdmann, A. L., Lanzoni,
G. M. D. M., & Higashi, G. D. C. (2013).
Myocardial revascularization: Strategies for coping
with the disease and the surgical process. Acta
Paulista de Enfermagem, 26(1), 8–13. https://doi.
org/10.1590/S0103-21002013000100003
Lam, S. C., Lee, D. T. F., & Yu, D. S. F. (2014).
Establishing cutoff values for the Simplified Barthel
Index in elderly adults in residential care homes.
Journal of the American Geriatrics Society, 5(62),
575–577. https://doi.org/10.1111/jgs.12716
Lee, E. T., & Wang, J. W. (2003). Identification of
prognostic factors related to survival time. In E. T.
Lee & J. W. Wang (Eds.), Statistical methods for
survival data analysis (pp. 243–280). Hoboken. NJ:
Wiley-Interscience.
Li, H. C., Chen, Y. S., Chiu, M. J., Fu, M. C.,
Huang, G. H., & Chen, C. C. H. (2015). Delirium,
subsyndromal delirium, and cognitive changes in
individuals undergoing elective coronary artery
bypass graft surgery. Journal of Cardiovascular
Nursing, 30(4), 340–345. https://doi.org/10.1097/
jcn.0000000000000170
Lopes, M. V. O., Silva, V. M., & Araujo, T. L. (2012).
Methods for establishing the accuracy of clinical
indicators in predicting nursing diagnoses.
International Journal Nursing Knowledge, 23, 134–139.
https://doi.org/10.1111/j.2047-3095.2012.01213.x
Magalhães, L. P., Figueiredo, M. J. O., Cintra, F. D.,
Saad, E. B., Kuniyoshi, R. R., Teixeira, R. A., …
Souza, O. F. (2016). II Brazilian Guidelines for
Atrial Fibrillation. Arquivos Brasileiros de Cardiologia,
106(Suppl. 2), 1–22. https://doi.org/10.5935/
abc.20160055
Mendonça, K. M. B., & Andrade, T. M. (2015).
Patient’s perception about coronary artery bypass
grafting. Brazilian Journal of Cardiovascular Surgery,
30(5), 544–551. https://doi.
org/10.5935/1678-9741.20150060
Minosso, J. S. M., Amendola, F., Alvarenga, M. R.
M., & de Campos Oliveira, M. A. (2010).
Validation of the Barthel Index in elderly patients
attended in outpatient clinics, in Brazil. Acta
Paulista de Enfermagem, 23(2), 218–223. https://doi.
org/10.1590/S0103-21002010000200011
Montrief, T., Koyfman, A., & Long, B. (2018).
Coronary artery bypass graft surgery complications:
A review for emergency clinicians. American Journal
of Emergency Medicine, 36(12), 2289–2297. https://
doi.org/10.1016/j.ajem.2018.09.014
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International
Prognostic Indicators of Delayed Surgical Recovery
Oliveira, F. S., Freitas, L. D. O., Rabelo-Silva, E. R.,
Costa, L. M., Kalil, R. A. K., & Moraes, M. A. P.
(2017). Predictors of mediastinitis risk after
coronary artery bypass surgery: Applicability of
score in 1.322 cases. Arquivos Brasileiros de
Cardiologia, 109(3), 207–212. https://doi.org/10.5935/
abc.20170119
Pepe, M. S. (2003). Incomplete data and imperfect
reference tests. In M. S. Pepe (Ed.). The statistical
evaluation of medical tests for classification and prediction
(pp. 549–553). New York, NY: Oxford University
Press.
Pimentel, M. F., Soares, M. J. F., Murad, J. A. J.,
Oliveira, M. A. B., Faria, F. L., Faveri, V. Z., …
Guido, R. C. (2017). Predictive factors of long-term
stay in the ICU after cardiac surgery: Logistic
CASUS score, serum bilirubin dosage and
extracorporeal circulation time. Brazilian Journal of
Cardiovascular Surgery, 32(5), 367–371. https://doi.
org/10.21470/1678-9741-2016-0072
Quintana, J. F., & Kalil, R. A. K. (2012). Heart
surgery: Feeling of patient before and after surgery.
Revista Psicologia Hospitalar, 10(2), 17–32.Retrieved
from http://pepsic.bvsalud.org/scielo.php?scrip
t=sci_arttext&pid=S1677-740920120002000
03&lng=pt&nrm=iso
Rembold, S. M., Santana, R. F., Lopes, M. V. O., &
Melo, U. G. (2020). Nursing diagnosis risk for
delayed surgical recovery (00246) in adult and
elderly: A case-control study. International Journal of
Nursing Knowledge, 31, 268–274. https://doi.
org/10.1111/2047-3095.12274
Reynolds, D., Duray, G. Z., Omar, R., Soejima, K.,
Neuzil, P., Zhang, S., … Ritter, P. (2016). A
leadless intracardiac transcatheter pacing system.
New England Journal of Medicine, 374(6), 533–541.
https://doi.org/10.1056/NEJMoa1511643
Rosso, R., & Viskin, S. (2020). Benign versus
malignant early repolarization patterns. In N.
El-Sherif (Ed.), Cardiac repolarization (pp. 277–283).
Cham, Switzerland: Springer. https://doi.
org/10.1007/978-3-030-22672-5_16
Rutjes, A. W. S., Reitsma, J. B., Coomarasamy, A.,
Khan, K. S., & Bossuyt, P. M. M. (2007).
Evaluation of diagnostic tests when there is no
gold standard: A review of methods. Health
Technology Assessment, 11(50), 3–9.
Santana, R. F., Pereira, S. K., Carmo, T. G., Freire, V.
E. C. S., Soares, T. D. S., Amaral, D. M., &
Vaqueiro, R. D. (2018). Effectiveness of a
telephone follow-up nursing intervention in
postsurgical patients. International Journal of Nursing
Practice, 24(4), e12648. https://doi.org/10.1111/
ijn.12648
437
Prognostic Indicators of Delayed Surgical Recovery
Shah, S., Vanclay, F., & Cooper, B. (1989). Improving
the sensitivity of the Barthel Index for stroke
rehabilitation. Journal of Clinical Epidemiology, 42,
703–709.
Silva, G. S., Sousa, A. G., Soares, D., Colósimo, F. C.,
& Piotto, R. F. (2013). Evaluation of the length of
hospital stay in cases of coronary artery bypass graft
by payer. Revista da Associação Médica Brasileira, 59(3),
248–253. https://doi.org/10.1016/j.ramb.2012.12.005
Silveira, C. R., Santos, M. B. K., Moraes, M. A. P.,
& Souza, E. N. (2016). Clinical outcomes of
patients undergoing cardiac surgery in a hospital in
northwest of Rio Grande do Sul. Revista de
Enfermagem UFSM, 6(1), 102–111. https://doi.
org/10.5902/2179769216467
Smith, C. D., & Custard, K. (2014). The experience
of family members of ICU patients who require
extensive monitoring: A qualitative study. Critical
Care Nursing Clinics of North America, 26(3), 377–
388. https://doi.org/10.1016/j.ccell.2014.04.004
Soares, G. M. T., Ferreira, D. C. S., Gonçalves, M. P.
C., Alves, T. G. S., David, F. L., Henriques, K. M.
438
Gomes do carmo et al.
C., & Riani, L. R. (2011). Prevalence of major
postoperative complications in cardiac surgery.
Revista Brasileira de Cardiologia, 24(3), 139–146.
Retrieved from http://www.onlineijcs.org/english/
sumario/24/pdf/v24n3a02.pdf
Souza, O. F., & Scanavacca, M. I. (2016). Cardiac
arrhythmias: Diagnosis and treatment. Rio de Janeiro,
Brazil: Rubio.
Zhou, X. H., Obuchowski, N. A., & McClish, D. K.
(2011). Methods for correcting imperfect gold
standard bias. In X. H. Zhou, N. A. Obuchowski,
& D. K. McClish (Eds.). Statistical methods in
diagnostic medicine (pp. 389–433). Hoboken, NJ:
Wiley.
Supporting Information
Additional supporting information may be found in
the online version of this article at the publisher’s
web site:
Fig S1
Table S1-S5
Journal of Nursing Scholarship, 2021; 53:4, 428–438.
© 2021 Sigma Theta Tau International