WHY MIGHT CLINICIANS IN MALAWI NOT OFFER HIV TESTING TO
THEIR PATIENTS?
Corey Lau MD1, MS; 2Adamson S. Muula MB BS, MPH; 3Humphreys Misiri, Msc;
3
Tilera Dzingomvera MB BS and Gregory Horwitz PhD, MS,
1
Current: Department of Epidemiology, UCLA School of Public Health, Los Angels, CA, USA
Prior: Department of Internal Medicine, Blantyre Adventist Hospital, Blantyre, Malawi
2
Department of Community Health, University of Malawi College of Medicine, Blantyre, Malawi
3
Out-Patient Clinic, Blantyre Adventist Hospital, Blantyre, Malawi
4
Vision Centre Laboratory, Salk Institute for Biological Studies, La Jolla, CA, USA
ABSTRACT
Context: HIV testing may be a cost-effective means of reducing transmission rates. Since
Malawi is severely impacted by HIV/AIDS and detection of infected persons is suboptimal, reasons clinicians might not offer HIV testing to their patients should be
identified.
Objective: Identify common reasons clinicians in Malawi might not offer HIV testing to
penitents
Design, Setting and Participants: Cross-sectional, descriptive, postal survey techniques
with telephone and fax follow-up were used to collect a census of clinicians in Malawi.
Basic demographic information and reasons one might not offer an HIV test were
solicited.
Main Outcome Measures: Proportions were calculated for each reason for not offering
HIV testing. Multiple logistic regression was used to determine whether responses
differed by demographic characteristics. A p-value of <0.05 was considered significant.
Results: 523 of 982 (53.26%) registered clinicians responded. Mean age of respondents
was 43.56 +/- 12.93. The five most common reasons for not offering HIV testing were:
inadequate training in HIV counseling (50.29%); perception that patient not ready to be
tested (49.52%); no indication for testing (35.56%); testing facilities unavailable
(35.37%); and insufficient time for testing (28.87%). Of six medical professional cadres/
categories, medical assistants were most likely to give the reason inadequate training in
HIV counseling (p=1.1200 x 10-8, _ coefficient = 1.2693 ± 0.2290).
Conclusions:
Though differences in reasons for not offering HIV tests amongst clinician groups were
small, medical assistants report lack of training in HIV counseling as a barrier to HIV
testing more than other groups. Thus development of one general program based on
common reasons identified in this study, with a subcomponent to educate medical
assistants on HIV counseling, could be effective in increasing HIV testing rates. Further
investigation of the identified reasons should be undertaken to facilitate program
development.
1
BACKGROUND:
HIV/ AIDS is a major global health concern. Over 42 million people were estimated to be
infected with the virus world- wide. Sub-Saharan Africa, where at least 29 million people
were infected at the end of 2002, is the region with highest prevalence.
It is estimated that at least fifteen percent of Malawi’s adult (15-49 years old) population
is HIV infectedii. The country is therefore amongst those that have been hardest hitiii,iv.
For example, HIV infection is responsible for the upsurge in the number of tuberculosis
(TB) cases, pneumonia, sepsis, Kaposi’s sarcoma and other cancers in Malawiv,vi,vii.
Furthermore, mortality is higher amongst children with HIV positive parentsviii.
Unfortunately, diagnosis of HIV usually late in the disease, when symptoms and AIDS
defining-illnesses are already present.
Benefits of early HIV diagnosis have become increasingly apparent with recent treatment
advancesix. A short time ago, diagnosis of HIV sero-postivity resulted only in
disheartenment due to lack of effective therapy. Even in the private sector where at least
some could afford drugs, particularly anti-retroviral therapy (ART), the outlook for HIV
positive persons was dismal. Presently, ART has been made more accessible at a number
of private hospitals, and even at selected public health institutions.x At end 2002, 1220
patients in Malawi’s public health sector were receiving ART (904 in Blantyre and
Lilongwe, 316 in Chiradzulu).xi The Ministry of Health and Population plans to increase
accessibility to ART through the Global Fund HIV/ AIDS Programs.xii Other measures to
improve ART availability are Church of Central Africa Presbyterian (CCAP) Blantyre
Synod Initiative through a donation from Pittsburgh Presbytery and involvement by nongovernmental organizations, for example Medicins Sans Frontieres (MSF).xiii
Studies suggest that worldwide screening levels for sexually transmitted diseases,
including HIV, are well below practice guidelines.xiv,xv For example, a survey of
members of the American Academy of Pediatrics found that 42.5% of pediatricians do
not counsel pregnant women or mothers of newborns on HIV screening. xvi Peters et al
reported that missed opportunities for perinatal HIV prevention contributed to more than
half of the cases of HIV infected infants.xvii However, implementation of an opt-out
prenatal HIV testing policy results in a dramatic increase in the number of females being
tested for HIV infection.xviii Furthermore, education of health care providers has been
shown to dramatically increase testing rates.xix,xx,xxi
Ethical dilemmas associated with HIV testing may prevent health care workers from
offering testing and may also inhibit patient acceptance of testing.xxii,xxiii In the United
States, New Zealand and Zimbabwe, surveys have revealed several reasons physicians
might not offer HIV testing to patients: perceived reluctance to be tested, lack of time,
knowledge deficit, insufficient support services, disinclination to inform a patient about
positive sero-status and fear of traumatizing patients.xxiv,xxv,xxvi However, HIV testing is a
cost-effective approach for preventing HIV infection.xxvii,xxviii In addition , routine
screening is likely more cost-effective than testing based on medical history.xxix
2
Knowledge about HIV status informs possible management options on the part of both
patients and health care workers. If clinicians do not offer HIV tests to their patients, the
benefits of HIV serostatus knowledge may be forfeited. Furthermore, benefits of early
diagnosis and treatment options, such as Highly Active Antiretroviral Theraphy
(HAART) for HIV, contrimoxazole prophylaxis against Pneumocystic carrinii
pneumonia and Fluconazole for Cryptococcal meningitis, continue to increase. Since
Malawi is so severely impacted by the HIV/ AIDS epidemic and identification of infected
persons is sub-optimal, measures must be taken to increase HIV testing rates. Patientclinician interactions are an invaluable and under-utilized opportunity for testing.
In this study, clinicians registered with the Medical Council of Malawi were surveyed to
identify reasons for not offering HIV testing to their patients. Differences in reasons
amongst clinician subgroups were also assessed. Elucidation of these reasons will
hopefully facilitate development of targeted interventions to increase rates of HIV testing,
thus improving management of sero-positive individuals and potentially decreasing rates
of HIV transmission.
METHODS:
Technique:
This was a cross-sectional, descriptive postal survey with telephone and fax follow-up.
Study Population:
All clinicians registered with the Medical Council of Malawi were recruited to participate
in this study. This group included specialists, general practitioners, dentists, clinical
officers and medical assistants from both public and private practices. Clinical Officers
and Medical Assistants comprise the largest groups of clinicians in Malawi.xxx
As the annually published Medical Council of Malawi Registry may be incomplete,
clinician names were also requested from district and mission hospitals and surveys
mailed to those people. Clinicians who were out of the country during the study period,
deceased or retired were excluded from the census. A total of 982 eligible clinicians were
identified.
Survey Instruments:
Instruments designed by the study authors were used to assess reasons for not offering
HIV testing. All surveys included instructions to print clearly and asked several basic
demographic questions. Following the demographic questions, participants were asked to
tick choices corresponding to and/ or write in their reasons for not offering HIV testing.
The survey is shown in Figure 1. An informed consent sheet accompanied each survey.
Return of the survey was considered indicative of willingness to participate in the study.
Ethical clearance was obtained from the College of Medicine Research and Ethics
Committee (COMREC).
3
Data Collection:
Postal survey techniques with telephone and fax follow up were used to collect a census.
Surveys were mailed to clinicians in May, 2003. A pre-addressed stamped envelope was
included to facilitate return by post. After one month non-responders were contacted by
telephone. Telephone numbers listed on the Medical Council of Malawi registry, with
district health offices and in the Malawi Telecommunications Limited directory were
used. When the clinician was unable to complete the survey via telephone, the survey was
faxed to the clinician and returned by fax to Blantyre Adventist Hospital.
Analysis:
Survey data was entered into a Microsoft Excel spreadsheet. To account for use of
different wording in the written response section, written responses from respondents
were grouped according to reason for purposes of analysis.
In order to eliminate census bias due to non-response, population weighting was used to
evaluate the effect of non-response related to profession.xxxi Proportions were calculated
for each listed reason for not offering HIV testing to patients. Multiple logistic regression
techniques were used to determine whether response differ by respondent age, gender,
citizenship or profession.xxxii The fitted was of the form: log [p/(1-p)] = _ 0 _1 (age) +_2
(gender) + _3 (citizenship) + _4 (specialist) +_5 (general practitioner) + _6 (medical
assistant) + _7 (dentist) + _8 (Other profession).
Each of the 16 reasons for not offering HIV testing was modeled independently. Clinical
officers comprised the largest profession proportion and were therefore used as the
referent group. S-plus statistical package was used for analysis.xxxiii Statistical
significance was assessed with likelihood ratio tests, A p-value of <0.05 was considered
statistically significant.
RESULTS:
523 of 982 clinicians (53.26%) responded to the postal survey. Demographic
characteristics of respondents are shown in Table 1. Average age respondents was 43.56
+/- 12.93 (mean +/- standard deviation); average age of the surveyed population was
46.90 +/- 12.28. Respondents were overwhelmingly males and Malawian citizens
(91.22% and 91.20%, respectively). 39.34% of respondents were in private; 40.84% were
trained in palliative care; 55.38% had HIV testing available through their practice;
22.46% worked at facilities offering HAART; 66.47% worked at facilities with a trained
HIV counselor; and 45.05% reported that their facility was associated with a home based
care program.
In Table 2, the reported professions of respondents and profession distribution for the
registry of the Medical Council of Malawi are shown. Distribution of professions was
different between respondents and the surveyed population; medical assistants were
under-represented and clinical officers were over-represented in the sample.
4
Reasons given for not offering HIV testing to patients are shown in Table 3. In addition
to the eleven suggested reasons, the following five were included based on written
responses: no facility available, no ARV available, non-conducive circumstances (e.g.
communication barriers, lack of privacy in the clinic), logistical difficulties (e.g. testing
center is too far, results take too long), and cost of testing and therapy. Note that the sum
of response percentages exceeds 100% because respondents were allowed to provide
multiple reasons.
Population weighting by profession did not affect relative order of reasons provided.
Population weighing by age was not performed because difference in mean ages between
the sample and surveyed population was 3.3446 years, which is relatively small.
Table 4 shows which factors were significant predictors for reasons for not offering HIV
testing. The first column lists reasons and factors (age or professional category) found to
be significantly correlated. P-values and fitted coefficients corresponding to each factor
are given.
DISCUSSION:
The most common reasons clinicians in Malawi give for not offering HIV testing to their
patients are that they are not adequately trained in HIV counseling (50.29%); perception
that the patient is not ready to be tested (49.52%); there is no indication for testing
(35.56%); testing facilities are unavailable (35.37%); and that there is insufficient time
available for testing (28.87%). Other reasons (patient personally known, non-conducive
circumstances first encounter with patient, poor prognosis, lack of ARV, guardian
present, logistical difficulties, patient age, not involved in clinical care, cost of testing of
therapy, and forgot to ask) were given by less than 20% of clinicians.
Amongst these five most common reasons, only inadequate training in HIV counseling
and testing, patient not ready to be tested and lack of testing facility were given
significantly more often by members of particular clinician subgroups. Specifically,
medical assistants were more likely to give the reasons not trained in HIV counseling
(1.2693 ± 0.2290, _ +/- standard error) and patient not ready to be tested (-0.5380 ±
0.2169); general practitioners were more likely to give the reasons no testing facility (08828 ± 0.3357) and patient not ready to be tested (0.6891 ± 0.3089); and those who
indicated other profession were more likely to give the reason lack of testing facility (1.04771± 0.5246). Thus, common reasons clinicians might not offer HIV testing to
patients differ slightly amongst professional groups. The strongest correlation was shown
between medical assistants, who constitute 46.33% of Malawi’s clinicians, and giving the
reason not trained in HIV counseling, the most common reason given by the all
clinicians (p=1.1200 x 108, _ = 1.2693 ± 0.2290).
Age is also significantly correlated with two of the five most common reasons for not
offering HIV testing to patients (Patient not ready to be tested, No time for testing).
However, the magnitude of the correlation of age with these reasons is small (0.02237 ±
0.007348 and –0.02736 ± 0.008384 respectively). Contribution of age to the odds ratio is
5
within the range 0.5330 (_ = -0.02736, age =23) to 0.1121 (_ = -0.02736, age = 80). In
contrast, the contribution of medical assistant to the odds ratio for the reason not trained
in HIV counseling is 3.5584 (_ = 1.2693). Thus, common reasons clinicians might not
offer HIV testing to patients are fairly homogenous across age groups.
Given that common reasons fro not offering HIV testing to patients differ slightly
amongst clinician groups, programs may need to develop profession-specific
interventions. It might be most practical to develop one collective intervention for all
clinicians in Malawi, rather than different interventions for different clinician groups, and
design and additional subcomponent to educate medical assistants on HIV counseling.
Although the reason lack of testing facilities was significantly correlated with profession,
clinician interventions will not increase accessibility of facilities. Thus no profession
specific program is necessary for this topic. However, all clinicians should be made
aware of available testing venues. Use of a common program would simplify the
development process and facilitate efficient use of resources.
While interventions to improve testing rates must target the most common reasons for not
offering HIV tests to patients, more detailed information on these reasons is necessary.
For example, since 50.29% of clinicians report that they are not adequately trained in
counseling, further education is needed in this area. Thus areas in which clinicians feel
their training is inadequate must be identified in order to improve education.
After inadequate training, the second and third most common reasons for not offering
HIV testing are the perception that patient is not ready and that there is no indication for
testing, respectively. In these cases, lowering the testing threshold might help minimize
the number of patients who remain untested. Though clinical judgment should not be
disregarded, it is possible that clinicians have an unnecessarily high threshold for
believing a patient ready or that testing is indicated. Considering that approximately one
in six adult Malawians is HIV infected, it is concerning that HIV testing may be viewed
as not indicated for persons seeking health care.
Lack of testing facility was the fourth most common reason listed for not offering testing
to patients. As test availability is improving in Malawi, Clinicians must be made aware of
options. At this time, the public health system offers HIV testing and ARV at Lilongwe
Central Hospital and Queen Elizabeth Central Hospital. District-based operations in
Thyolo and Chiradzulu, as well as select private institutions, also provide these
services.xxxiv Thus, a clinician whose facility does not have its own testing center might
send blood samples to a nearby testing center for analysis or refer patients to the other
center. Efforts are also being made to establish additional testing centers, which should
facilitate improved testing rates. Furthermore, alternative methods, such as saliva tests,
might eventually be available for screening.xxxv Though confirmatory testing would still
be necessary, use of alternative methods could simplify the screening process.xxxvi
Lack of time, reported by 28.87% of respondents, was the fifth most common reason for
not offering HIV tests to patients. Time constraints are a universal problem. Besides
teaching clinicians how to improve their time management, clinicians might also be
6
advised to use ancillary services. Use of ancillary services, such as in-house counselors or
voluntary counseling and testing centers, saves the clinician time during the medical visit
while simultaneously providing a venue for the patient to undergo testing.
Information from this study could be used to develop interventions that may improve
HIV testing rates in Malawi. However, this study is limited by several factors. Given the
53.26% response rate, the sample may not be representative of the population of all
clinicians in Malawi. Non-random non-response could bias results. Fortunately,
distribution of age in the sample and surveyed population were similar, and population
weighting by profession did not change the relative frequency of responses. Thus, results
are likely generalizable to Malawi’s clinician population.
The surveyed population was limited to clinicians listed with the Medical Council of
Malawi plus those known by the investigators. It is thus possible that some clinicians
were not surveyed. In addition, some respondents did not provide answers to all
questions. Responses for surveys that did not contain all factors in a particular model
could not be considered in the model fit. Since the maximum number of surveys excluded
for a particular model was 23 out of 523 (4.40%), exclusion of these is unlikely to
significantly impact results.
In conclusion, efforts to improve HIV testing rates in Malawi should include measures to
increase clinician offering of HIV tests during patient encounters. Such targeted
interventions must take into account reasons why clinicians do not offer testing to their
patients. In this study, several common reasons for not offering HIV tests were identified.
Based on these results, it seems most appropriate to develop one general clinician’s
intervention with an additional component to reduce medical assistants on HIV
counseling. However, further characterization of reasons for not offering HIV tests
should be performed to facilitate effective intervention development.
Figure 1: Survey sent to all clinicians
Clinician Survey
Please tick or print your responses clearly.
Number:
What is your age in years? _______________________
What is your gender?
____Male
____Female
Are you a citizen of Malawi? ____Yes
____No
How would you describe your practice?
____General Practice
____Internal Medicine
___Obstetrics/Gynecology
____Pediatrics
____Psychiatry
____Surgery
___Medical Assistant
____Clinical Officer
____Dentistry
___Other (please specify):________
Are you in private practice?
____Yes
____No
Were you trained in palliative care?
____Yes
____No
Is HIV testing available at your practice?
___Yes
____No
___ Not sure
Does your facility offer Anti-Retroviral Therapy? ____Yes ____No Not sure
7
Is your facility associated with an HIV home-based care program? _Yes _ No
_Not sure
Please tick any / all reasons you may not have been able to offer an HIV test:
_Not enough time
_ A guardian is present
_I am not trained in HIV counseling
_The patient is not ready to be tested
_First encounter with the patient
_I personally know the patient
_HIV testing is not indicated
_HIV has no cure/ Prognosis is poor
_I am not involved in clinical care
_I forgot to ask
_Age of the patient
Please list any other reasons that have prevented you from offering an HIV test:
________________________________________________________________________
__________________________________________________________
Please return your complete survey in the envelope provided. Thank you.
Table 1: Demographic factors respondents
FACTOR
Age – respondents
Age – population surveyed
Gender
Citizenship
Private practice
Palliative Care Training
HIV testing Available
Anti-retroviral Therapy Available
Trained HIV Counselor Present
Home Care Available
VALUE
46.90 ±12.23
43.56 ±12.93
478 (91.22%) Male
477 (91.20%) Malawian
203 (39.34%) Private
194 (40.84%) Trained
288 (55.38%) Available
117 (22.46%) Available
343 (66.47%) Trained
232 (45.05%) Available
SAMPLE SIZE
512
771
524
523
516
475
520
521
515
515
Table 1 shows demographic factors of survey respondents. For Age, the mean +/Standard deviation is reported. The age value for the surveyed population takes into
account respondents and non-respondents. For other factors, number of clinicians
reporting a particular characteristic are given. Values in parentheses show the
corresponding percentage of respondent reporting characteristic. Sample sizes vary
because some respondents did not provide information on some factors.
Table 2: Professional categories of respondents
PROFESSIONAL
Clinical Officer
Medical Assistant
General Practitioner
Specialist*
RESPONDENTS
195 (37.28%)
177 (33.84%)
75 (14.34%)
39 (7.45%)
POPULATION
294 (29.94%)
455 (46.33%)
146 (14.87%)
62 (6.31%)
8
Dentist
Other
7 (1.34%)
30 (5.74%)
25 (2.55%)
0 (0.00%)
Table 2 shows the distribution of professions amongst respondents and the entire
surveyed population. Numbers reporting particular professions are given. In parentheses,
corresponding percentages of clinicians in that category reporting the specific profession
are shown.
*Specialist includes Surgery, Obstetrics-gynecology, Internal Medicine, Surgery and
Psychiatry
Table 3: Reasons fro not offering HIV testing to patients
Response
Not a Trained Counselor
Patient Not Ready
No Indication for Testing
No Facility Available
Not Enough Time
Patient Personally Known
Circumstances
First Clinical Encounter
Poor Prognosis
No HAART Available
Guardian Present
Logistics of Testing
Patient Age
Not Involved in Care
Cost of Testing/ Therapy
Forgot to Ask
Number Reporting
263 (50.29%)
259 (49.52%)
186 (35.56%)
185 (35.37%)
151 (28.87%)
92 (17.59%)
83 (15.87%)
61 (11.66%)
56 (10.71%)
44 (8.41%)
39 (7.46%)
29 (5.54%)
28 5.34%)
25 (4.78%)
25 (4.78%)
17 (3.25%)
Table 3 shows the distribution of responses for the entire sample. Total sample size is
523. Numbers in parentheses are percentages of total sample giving a particular response.
These values are not population weighed.
Table 4: Magnitude of effect for significant correlations
Correlated Reason / Factor
Not A Trained Counselor
Medical Assistant
Patient Not Ready
Age
Patient Not Ready
Medical Assistant
Patient Not Ready
General Practitioner
P value
1.12oo x 10-8
_ Coefficient
1.2693 ± 0.2290
0.01294
-0.02237 ± 0.007348
0.01279
-0.5380 ± 0.2169
0.02299
0.6891 ± 0.3089
9
No Testing Facility
General Practitioner
No Testing Facility
Other Profession
Not Enough Time
Age
Not Involved in Care
Age
Not Involved in Care
Other Profession
Patient Personally Known
Medical Assistant
No ARV Available
Medical Assistant
No ARV Available
Other Profession
Circumstances
Specialist
Circumstances
General Practitioner
Circumstances
Other Profession
Logistics
Specialist
0.006273
-8828 ± 0.3357
0.03231
-1.04771 ± 0.5246
0.006987
-0.02736 ± 0.008384
0.005041
0.06986 ± 0.01636
0.04422
1.7123 ± 0.7741
0.03194
-061544 ± 0.2916
0.01246
-1.0146 ± 0.4292
0.009260
-7.2812 ± 11.3822
0.02965
-1.3265 ± 0.6822
0.03531
-0.8844 ± 0.4466
0.04125
-1.3704 ± 0.7735
0.01850
-6.9645 ± 9.7318
Table 4 shows the p values and _ coefficients for factors significantly related to reporting
a reason for not offering HIV testing. In the “correlated reason /factors column”, the
reason for not offering HIV tested is listed first; factor found to be predictive of listing
that reason is listed second. The “_ Coefficient” column shows the _ Coefficient +/standard error for the designated factor.
ACKNOWLEDGEMENTS:
The authors thank the National Research Council of Malawi for project funding, Blantyre
Adventist Hospital for logistical support; and Andrew Likaka, Malangizo Mbewe.
Lumbani Munthali and Sandress Msuku for data entry.
10
REFERENCES
UNAIDS, WHO, AIDS Epidemic Update 2002. UNAIDS, Geneva Switzerland, 2002.
UNAIDS, WHO, Epidemiological Fact Sheet: Malawi, UNAIDS, Geneva Switzerland,
2002.
Chirwa I. AIDS Epidemic in Malawi: shaking cultural foundations. Network 1993; 13
(4): 21-31.
Cuddington JT & Hancock JD. Assessing the impact of AIDS on the growth path of the
Malawian economy. J. Dev Econ 1994: 43 (2): 362-368.
Chimzizi RB, Harries AD, Hargreaves NJ, Kwanjana JH, Salaniponi FM. Care of HIV
complication in patients receiving anti-tuberculous treatment in hospitals in
Malawi. Int J. Tuberc Lung Dis 2001; 5(100): 979-981.
Glynn JR, Warndoff DK, Malema SS, et al. Tuberculosis: associates with HIV and
socioeconomic status in rural Malawi. Trans R. Soc Trop Med Hyg 2000; 94(5):
500-503.
Graham SM, Mtitimila El, Kamanga HS, Walsh AL, Hart CA, Molyneux ME. Clinical
presentation and outcome pf Pneumocystis carinii pneumonia in Malawian
Children. Lancet 2000; 355 (9201): 369-373.
Crampin AC, Floyd S, Glynn JR, et al. The long-term impact of HIV and orphanhood on
the mortality and physical well-being of children in rural Malawi. AIDS 2003; 17
(3): 3 89-97
Vaidiserri RO, Janseen RS, Buehler JW, Fleming PL. The context of HIV/AIDS
surveillance. J Acquir Immune Defic Syndr 2000;25 (Supp 2): S97-104.
Nachega J, Antiretroviral Treatment in Developing Countries. The Hopkins HIV Report
2002; 14 (5): 10-11.
Ministry of Health and Population and National AIDS Commission. Treatment of AIDS.
Guidelines for the use of antiretroviral therapy in Malawi, 1st ed. Ministry of
Health and Population 2003.
National AIDS Commission. Global Fund Grant Award to Malawi. Daily Times Nov 9,
2002: 8.
Jamali P. Church pledges Anti-Retroviral Drugs. Battle Against AIDS Rages On. The
Weekend Nation, November 9-10, 2002: 11
St Lawrence JS, Montano DE, Kaspryzk D, Phillips WR, Armstrong K, Leichliter JS.
STD screening, testing, case reporting, ad clinical and partner notification
practices: a national survey of US physicians. American Journal of Public Health
2002; 92 (11): 1784-1788.
Centres for Disease Control and Prevention, Revised guidelines for HIV counseling,
testing and referral. MMWR Recomm Rep 2001; 50 (RR-19): 1-57.
Kine MW, O’Connor KG. Disparity between pediatricians’ knowledge and practices
regarding performing HIV counseling and testing. Pediatrics 2003; 112 (5): e367.
Peters V, Liu KL, Doninguez K, et al. Missed opportunities for perinatal HIV prevention
among HIV-exposed infants in 1996-2000, pediatric of HIV disease cohort.
Pediatrics 2003; 111 (5 part 2): 1186-91.
Jayaraman GC, Preiksaitis JK, Larke B. Mandatory reporting of HIV infection and optout prenatal screening for HIV infection: effect on testing rates. CMAJ 2003; 168
(6): 679-82.
11
Grimes RM, Courtney CC, Vindekilde J. A collaborative program between a school of
public health and a local health department to increase HIV testing of pregnant
women. Public Health Rep 2001; 116 (6): 585-589.
Lalonde B, Uldall KK, Huba GJ, et al. Impact of HIV/AIDS education on health care
provider practice: results from nine grantees of the Special Projects of National
Significance Program. Eval Health Prof 2002; 25 (3): 302-320.
Wolf MS, Mitchell CG. Preparing social workers to address HIV/AIDS prevention and
detection: implications for professional training and education. J. Community
Health 2002; 27 (3): 165-180.
Temmerman M, Ndinya-Achola J, Amabni P, Piot P. The right not to know HIV-test
results. Lancet North Am Ed 1995;345(8955): 969-970.
Makura ZGC. Competent inadequacy. Centr Afr Med J 1991; 37(2): 67-68.
Chambers ST, Heckert KA, Bagshaw S, Ussher J, Birch M, Wilson MA. Maternity care
providers’ attitudes and practices concerning HIV testing during pregnancy;
results of a survey of the Canterbury and upper South Island region. NZ Med J
2001; 114 (1144): 513-516.
Gibney L, Wade S, Madzime S, Mbizvo M. HIV testing practices of Zimbabwean
physicians and their perspectives on the future use of rapid on-site tests. AIDS
Care 1999; 11 (6): 663-673.
Troccoli K, Pollard H 3rd, McMahon M, Foust E, Erickson K, Schulkin J. Human
immunodeficiency virus counseling and testing practices among North Carolina
providers. Obstet Gynecol 2002; 100 (3): 420-427.
Sweat M, Gregorich S, Sangiwa G, et al. Cost effictiveness of voluntary HIV-1 testing in
reducing sexual transmission of HIV-1 in Kenya and Tanzania. Lancet 2000; 356
(9224): 113-121.
The Voluntary HIV-1 Counseling and Testing Efficacy Study Group. Efficacy of
voluntary HIV-1 counselling and testing in individuals and couples in Kenya,
Tanzania and Trinidad: a randomized trial. Lancet 2000; 356(9224) 103-112.
Gibney L, Wade S. Madzime S, Mbizvo M. HIV testing practices of Zimbabwean
physicians and their perspectives on the future use of rapid on-site tests. AIDS
Care 1999; 11 (6): 663-673.
Muula AS, Broadhead RL. The first decade of the Malawi College of Medicine: a critical
appraisal. Tropic Medicine and International Health 2001;6 (2): 155-9.
Groves RM and Couper MP. Nonresponse in Household Interview Survey. John Wiley
and Sons 1998, New York.
Rosner B. Fundamentals of Biostatistics. 5th edition. Pacific Grove (CA): Duxbury; 2000.
Data Analysis Division. S-PLUS 6.0 for UNIX. Seattle (WA): Mathsoft Inc; 2000.
Kemp J, Aitken JM, LeGrand S, Mwale B. Equity in health sector responses to
HIV/AIDS in Malawi. Regional Network for Equity in Health in Southern Africa.
Harare (Zimbabwe): EQUINET; 2003.
Luo N, Kasolo F, Ngwenya BK, Du Pont HL, Zumla A. Use of saliva as an alternative to
serum for HIV screening in Africa. South African Medical Journal 1995; 85(3):
156-7.
Samuel NM, Chandrasekaran A, Paul SA. Detection of antibodies of HIV-1 in serum
and saliva. J Assoc Physicians India 1997; 45 (4): 280-2.
12