AIDS and Behavior

, Volume 17, Issue 6, pp 2269–2282

How Informed are Clients Who Consent? A Mixed-Method Evaluation of Comprehension Among Clients of Male Circumcision Services in Zambia and Swaziland

Authors

    • Population Council
  • Louis Apicella
    • Population Council
  • Katie D. Schenk
    • Population Council
  • Meredith Sheehy
    • Population Council
  • Paul C. Hewett
    • Population Council
Original Paper

DOI: 10.1007/s10461-013-0424-1

Cite this article as:
Friedland, B.A., Apicella, L., Schenk, K.D. et al. AIDS Behav (2013) 17: 2269. doi:10.1007/s10461-013-0424-1

Abstract

Comprehension is fundamental for informed consent—an individual’s right to choose a medical procedure, such as male circumcision (MC). Because optimal benefits depend on post-surgical behaviors, comprehension is particularly critical for MC programs. We evaluated clients’ comprehension of MC’s risks and benefits, wound care instructions, and risk reduction post-MC using a true/false test (n = 1181) and 92 semi-structured interviews (SSIs) in Zambia and Swaziland. Most participants (89 % Zambia, 93 % Swaziland) passed the true/false test, although adolescents scored lower (significantly so in Swaziland) than adults and one-third (including nearly half of adolescents in Zambia) said MC has no risks. SSIs indicated confusion between “risk” of adverse surgical outcomes and reduced “risk” of HIV; most respondents acknowledged the 6 week abstinence period post-MC, yet few said resuming sex early increases HIV risk. Providers should distinguish between surgical “risks” and reduced HIV “risk,” and emphasize that HIV risk increases with sex before complete healing.

Keywords

Male circumcisionComprehensionInformed consentRisksRisk behaviorVMMC programs

Introduction

Three randomized controlled trials demonstrated that male circumcision (MC) reduces the risk of HIV infection in heterosexual men by approximately 60 % [13]. The trial results led the World Health Organization/The Joint United Nations Programme on HIV/AIDS (WHO/UNAIDS) to recommend the rapid scale-up of voluntary medical male circumcision (VMMC) services in regions with high HIV prevalence and low rates of circumcision [4]. Swaziland and Zambia are two of thirteen priority countries identified with high HIV prevalence (25.9 % Swaziland, 13.5 % Zambia) [5] and low MC rates (8.3 % Swaziland, 12.8 %, Zambia) [6, 7]. In 2009, the governments of Swaziland and Zambia introduced national policies for scaling up VMMC services with the goal of circumcising 80 % of 15–49-year-old males [8]. Although rapid scale-up of MC is imperative to the success of population level benefits [911], health activists, human rights advocates, researchers, and policy makers have emphasized the importance of adopting a human rights based approach to ensure that MC services are carried out safely, without coercion or discrimination, and with an effective informed consent process [12].

Informed consent (IC) is the fundamental right of an individual to decide whether or not to undergo a medical treatment or procedure. In contrast to medical research, for which specific guidelines and regulations exist [1316], IC in clinical practice has evolved primarily from Western case law [17, 18]. Experts on IC in clinical practice suggest that, to preserve every patient’s dignity and autonomy, health professionals have a duty to: (1) provide accurate and understandable information to the patient (or client) about both the benefits and potential risks of a procedure; (2) assess the patient’s understanding of the information; (3) assess the capacity of the patient to make the necessary decision; and (4) assure the patient has the freedom to choose without coercion or manipulation [19].

To ensure IC for circumcision, UNAIDS emphasizes the importance of promoting VMMC within a comprehensive HIV prevention framework [12]. Because MC is only partially effective against HIV and other sexually transmitted infections (STIs), provides no direct protection against HIV for women, and because MC’s population level benefit can be offset by changes in individual risk behavior [10, 11, 2025], policy makers and providers are concerned about the initiation of unsafe behaviors after circumcision (decreased condom use, increased number of sexual partners) or “risk compensation” [12]. Although empirical evidence of risk compensation is scant [13, 24, 26], several recent studies indicate that a notable proportion of men resume sex before the end of the recommended six-week healing period [1, 9, 2729], which can increase HIV risk for recently circumcised men and their partners [25, 27, 30]. Consequently, the IC process for VMMC must include a thorough explanation of the surgical procedure, including potential risks and benefits; information about the limitations of MC (partial protection against HIV and STIs, no direct HIV protection for women) and the need for continued risk reduction strategies; specific instructions for wound care; and concrete information about the required post-procedure abstinence period [31].

Despite widespread agreement about the importance of IC for VMMC, implementing an effective IC process is often challenging, particularly in countries characterized by limited resources and low literacy, including many places most affected by the HIV epidemic [32, 33]. Therefore, it is important to evaluate the MC IC process and client comprehension, in particular, to ensure that individual human rights are protected and benefits of MC are maximized as VMMC services are scaled up [12].

This paper reports on the first evaluation of the IC process for VMMC services in Swaziland and Zambia. The evaluation was conducted in the context of VMMC programs that were going to scale in each country. The objectives of the study were to: (1) assess male clients’ understanding of key concepts in the IC process; (2) examine social norms and practices regarding IC for adolescents; and (3) investigate how MC clients who had recently undergone circumcision felt their experiences matched their expectations based on pre-surgical counseling. This paper emphasizes findings related to the first objective, which was to assess MC clients’ comprehension of risks and benefits of the procedure, the need for post-procedure abstinence, and strategies for continued HIV risk reduction.

Methods

Study Design and Procedures

This study involved both quantitative and qualitative research methods that were incorporated into standard procedures for clients undergoing MC at participating sites (Fig. 1). The quantitative portion of the study used a post-test design in which an interviewer who was not part of the MC service delivery team administered a ten-question, true/false comprehension test to study participants after MC counseling (group and/or one-on-one) and voluntary HIV counseling and testing, prior to consenting for surgery. The post-test design was chosen because, regardless of when clients learn about MC (before or after coming to the clinic), their understanding of key concepts at the point in time when they consent to MC surgery is most relevant. The comprehension test was also developed with the potential to be integrated within the VMMC service delivery process. As there is no standard level of comprehension for IC, the research team, in consultation with service providers, set the benchmark as being able to correctly respond to at least eight out of ten questions on the true/false test. A null hypothesis was set such that at least 90 % of study participants could achieve a score of 80 %. Prior to taking the comprehension test, clients were asked to provide demographic data and answer several structured questions about their decision making process. Brief numeracy and literacy tests were also administered.
https://static-content.springer.com/image/art%3A10.1007%2Fs10461-013-0424-1/MediaObjects/10461_2013_424_Fig1_HTML.gif
Fig. 1

Study design integrated into standard procedures for clients undergoing male circumcision

The qualitative portion of the study involved semi-structured interviews (SSIs) with a different set of clients (adults and adolescents) one week post-MC. SSIs explored clients’ motivations for MC and comfort with the decision; comprehension of key concepts in the MC IC process—risks and benefits, instructions for wound care, partial protection, post-procedure abstinence, and the need for continued HIV risk reduction post-MC; and how experiences matched expectations, with an emphasis on how the consent process could be improved for future clients. At the beginning of the interview, clients were asked to describe themselves with respect to their age, marital status, whether they currently had a steady partner, and the highest level of education they had attained. Interviews were conducted by researchers not affiliated with the MC sites seven days after the MC procedure, usually at the clinic during the seven-day post-operative follow-up visit, or in the participant’s home, if preferred (Fig. 1).

The comprehension test and the interview guide for the SSIs were developed based on existing client brochures and counseling messages used by the participating sites. The study was implemented first in Zambia and then Swaziland; several true/false questions were modified for Swaziland based on the results from Zambia. All instruments were translated into local languages and pre-tested prior to study implementation. Study participants could elect to take the comprehension test in Bemba, Nyanja or English in Zambia, and Siswati or English in Swaziland. Comprehension test responses were entered by trained enumerators directly into hand-held computers programmed with Perseus Survey Solutions 7 software (Vovice, Herndon, Virginia, USA). The SSIs were conducted in the language or mixture of languages that enabled the most comfortable discussion between the respondent and the interviewer. SSIs were audio-recorded, transcribed, and translated into English.

Study Population and Sample Size

The study was conducted among a convenience sample of adults (18 years or older) and adolescents (13–17 years old) at 12 sites in Lusaka, Zambia and 11 sites in Swaziland in a mix of urban (primarily) and rural settings. All participants were recruited for the study on the day of the MC procedure, whether participating in the quantitative comprehension test or the SSIs. Only adolescents whose parent or legal guardian was present on the day of the MC procedure were eligible to participate. For the quantitative comprehension test, a sample size of 86 participants per cohort (adults and adolescents) was determined to be sufficient to achieve 80 % power and an alpha of 0.05 to evaluate the null hypothesis that 90 % of participants could pass the test. The initial sample size was increased to 100 per cohort to ensure sufficient data for analysis to account for potential line item missing cases. In Zambia, the target sample size was further increased to enable comparisons between adults and adolescents, static versus outreach clinics, and high- (15 MC surgeries per day or more) versus low-volume (less than 15 MCs per day) clinics, yielding a total sample of 400 study participants (300 adults, 100 adolescents). In Swaziland, the target sample size was increased to 1,125 participants (900 adults, 225 adolescents) for multiple secondary analyses by age, clinic type (rural/urban), site volume, and affiliation (NGO or government). For the SSIs, the aim was to interview a sufficient number of respondents (estimated to be 20–30 respondents from each age group in each country) until response content saturation was reached [34].

Data Analysis

Quantitative: Comprehension Test

Study participants received one point for each correct response on the ten-question true/false test and scores ranged from zero to 100 % correct. Participants who answered eight or more questions correctly were coded with a “1” (pass) and those who did not with a “0” (fail). A one-sample, one-sided t-test was used to assess whether the benchmark of 90 % of participants passing was met. Pearson’s Chi square was used to test responses to individual questions and the overall pass/fail indicator. Bivariate statistics and multiple logistic regression were used to examine relationships between background factors and comprehension scores. Stata version 10.1SE (StataCorp LP, College Station, Texas, USA) was used for all analyses.

Qualitative: Semi-Structured Interviews

Transcripts were imported into ATLAS.ti Version 5.2 (ATLAS.ti GmbH, Berlin [“City West”] Germany) to facilitate analysis. A code book was developed by three qualitative researchers with key domains of interest outlined prior to the study. Additional codes were added based on themes that emerged during the interviews. To ensure quality and inter-coder reliability, each transcript was coded by at least two team members and 10 % of the transcripts were coded by all three analysts. A modified grounded theory and framework analysis approach was used for analysis, enabling the systematic discovery of categories, themes, and patterns that emerged from the data [35, 36].

Ethical Considerations

The protocol was reviewed and approved by the Institutional Review Board of the Population Council, the University of Zambia Biomedical Research Ethics Committee, the Scientific and Ethics Committee of Swaziland, and the Zambian Ministry of Health. All participants gave written IC to take part in the study and were reassured that their health services and future medical care would not be affected if they declined participation. All interviews were conducted in private. For adolescent participants, parental/guardian consent was sought first, followed by adolescent assent. Research activities involving adolescents were informed by Ethical Approaches to Gathering Information from Children and Adolescents in International Settings: Guidelines and Resources [37].

Results

Comprehension Test

In Zambia, a total of 426 MC participants (311 adults, 115 adolescents) completed the comprehension test, however, 198 interviews were excluded from the analysis after it was determined that these participants might have been given the test before completing all counseling procedures. The reduced sample (159 adults, 69 adolescents) was sufficient for answering the primary research question; however, comparisons between high- and low-volume clinics and between static and outreach sites were no longer feasible. In Swaziland, 953 participants (756 adults, 197 adolescents) completed the comprehension test.

Characteristics of Clients Completing the Comprehension Test

As shown in Table 1, the majority of adults in both countries completed at least a secondary education and more than 90 % of all participants passed both numeracy and literacy tests. Study participants averaged 22.6 years of age in Zambia (25.7 years, adults; 15.4 years, adolescents) and 23.4 years in Swaziland (25.7 years, adults; 14.8 years, adolescents). The majority of participants had never been married. In Swaziland, nearly all participants (96 %) chose to take the test in Siswati, whereas in Zambia, the majority (73 %) chose to take it in English.
Table 1

Characteristics of clients participating in the true/false comprehension test, by country and by age group

 

Zambia

Swaziland

Adolescents (n = 69)

%

Adults (n = 159)

%

All (n = 228)

%

Adolescents (n = 197)

%

Adults (n = 756)

%

All (n = 953)

%

Mean age in years (range)

15.4

(13–17)

25.7

(18–58)

22.6

(13–58)

14.8

(13–17)

25.7

(18–62)

23.4

(13–62)

Marital status

 Single

100

73

81

100

78

82

 Currently marrieda

0

24

17

0

19

16

 Divorced, widowed, separated

0

3

2

0

2

1

Highest education completed

 None

2

0

<1

1

2

2

 Primary (grades 1–7)

17

6

9

59

14

23

 Junior secondaryb

41

13

22

36

22

25

 Senior secondaryc

39

37

38

4

41

34

 Post-secondary

2

44

31

0

20

16

Most comfortable spoken language

 Bemba

19

11

14

n/a

n/a

n/a

 English

29

50

44

2

3

3

 Nyanja

48

30

36

n/a

n/a

n/a

 Siswati

n/a

n/a

n/a

99

96

96

 Other

4

8

7

0

1

<1

Passed literacy test (in preferred language)

93

91

92

90

92

92

Passed numeracy test

100

99

99

99

98

98

Survey language chosen

 Bemba

6

4

4

n/a

n/a

n/a

 English

62

77

73

2

6

5

 Nyanja

32

19

23

n/a

n/a

n/a

 Siswati

n/a

n/a

n/a

98

94

95

Clinic type

 Fixed

70

74

73

97

79

82

 Outreach

30

36

27

3

21

17

Clinic location

 Rural

0

0

0

3

9

8

 Urban

100

100

100

97

91

92

Clinic volumed

 High (15+ MC per day)

72

52

58

n/a

n/a

n/a

 Low (<15 MC per day)

28

48

42

n/a

n/a

n/a

Type of counselinge

 Group, 1-on-1, and HIV counseling & testing (CT)

14

31

28

81

60

64

 Group and CT (no 1-on-1)e

n/a

n/a

n/a

12

5

6

 Group and 1-on-1 (no CT)

14

22

21

<1

2

1

 1-on-1 and CT (no group)

71

47

51

4

30

25

 1-on-1 only

0

0

0

<1

1

1

 CT onlye

n/a

n/a

n/a

2

1

1

 Group onlye

n/a

n/a

n/a

1

1

1

 Counseled in most comfortable language

51

60

57

n/a

n/a

n/a

 Group counseling in most comfortable languagef

n/a

n/a

n/a

90

86

87

 1-on-1 counseling in most comfortable languagef

n/a

n/a

n/a

98

84

87

Source of information on MCg

 Friend

20

63

50

20

26

25

 Parents

61

8

24

20

4

7

 Media

15

50

39

24

44

40

 Clinic outreach

11

28

23

4

13

10

 School

n/a

n/a

n/a

61

11

22

 Otherh

20

35

31

21

19

19

Recruited by interpersonal communication (IPC) agent/MC mobilizer

7

13

11

7

30

24

Decision comforti

 Very comfortable

46

70

63

48

71

66

 Moderately comfortable

10

18

16

35

23

25

 Somewhat uncomfortable

28

10

15

17

6

8

 Very uncomfortable

16

3

7

1

1

1

Felt he could change his mind about decision

76

63

67

26

73

64

n/a data not available/collected (or relevant, in the case of certain languages) in specified country

aIncludes currently married or living as married

bIn Swaziland, this category is called secondary (Form 1–3)

cIn Swaziland, this category is called high school (Form 4–5)

dClinic volume was not assessed in Swaziland

eIn Zambia, clients who reported not having 1-on-1 counseling were excluded from analysis because of concerns that the comprehension test may have been administered prior to 1-on-1 counseling

fIn Zambia, clients were asked about counseling language, overall, whereas in Swaziland they were asked specifically about language for group and 1-on-1 counseling

gClients could choose more than one option so responses may not add up to 100 %

hOther includes sibling, spouse, other family member, church, community elders/leaders, government

iMay not add up to 100 % due to rounding

As shown in Table 1, 56 % of participants in Zambia and 64 % in Swaziland reported receiving the recommended counseling package (group and one-on-one MC counseling, plus counseling and testing for HIV) [31]. In Zambia, 57 % reported being counseled in their most comfortable language whereas in Swaziland, where more detailed questions were asked, 87 % reported receiving both group and one-on-one counseling in their most comfortable language.

In Zambia, the majority of adults (63 %) first learned about MC from friends whereas the majority of adolescents (61 %) first heard about MC from their parents (Table 1). In Swaziland, where more detailed questions were asked about information sources, the primary source for adolescents was school (61 %), whereas adults reported first hearing about MC from media (44 %) and friends (26 %). Most clients (73 % Zambia, 79 % Swaziland) reported having made the decision to be circumcised prior to coming to the clinic on the day of the MC surgery. In both countries, the majority of clients reported being “very comfortable” with the decision to undergo MC (63 % Zambia, 66 % Swaziland); however, more adults (70 % Zambia, 71 % Swaziland) than adolescents (46 % Zambia, 48 % Swaziland) said they were “very comfortable” with the decision. A similar proportion of clients in both countries reported feeling like they could change their minds (67 % Zambia, 64 % Swaziland), however, there were marked differences within and between the two countries; in Zambia, fewer adults (63 %) than adolescents (76 %) felt they could change their minds, yet in Swaziland, 73 % of adults versus only 26 % of adolescents felt they could change their minds.

Comprehension Test Results

Most participants passed the comprehension test, answering at least eight out of ten questions correctly (Table 2). In Zambia, 89 % passed, which was not significantly lower than the 90 % benchmark (p = 0.51). In Swaziland, 93 % passed, which was significantly higher than the 90 % hypothesized (p < 0.001). In both countries, the question with the most incorrect responses was about risks involved in the MC surgery (Table 2, Question 3). In Zambia, the question was “There are no risks in MC surgery (FALSE),” which 33 % of participants answered incorrectly. However, nearly 95 % of Zambian clients correctly answered the question “It is possible to have pain, swelling, and bleeding after the MC procedure, but resting for 1–2 days after the surgery will help the wound heal” (Table 2, Question 2). In Swaziland, the question about risks (Table 2, Question 3) was reframed as “Risks of the MC procedure include pain, swelling, or bleeding during or after surgery [TRUE],” which 13 % of clients incorrectly answered. Questions about the pre-surgical anesthesia injection (Table 2, Question 1) and wound healing (Table 2, Question 2), yielded the most correct responses (nearly 100 % Zambia, 93 % Swaziland, and 95 % Zambia, 97 % Swaziland, respectively). Participants also scored well on questions related to safer sex practices post-MC (Table 2, Question 4, 6, and 9); for example, 94 % of participants in Zambia and 97 % in Swaziland correctly answered that circumcised men still need to use condoms. Similarly, 94 % of participants in Zambia and 96 % in Swaziland said that circumcised men still need to reduce their number of sexual partners.
Table 2

Results of the true/false comprehension test, by country and by age group

 

Zambia

Swaziland

Adolescents (n = 69)

% (95 % CI)

Adults (n = 159)

% (95 % CI)

All (n = 228)

% (95 % CI)

Pearson χ2

Pr=

Adolescents (n = 197)

% (95 % CI)

Adults (n = 756)

% (95 % CI)

All (n = 953)

% (95 % CI)

Pearson χ2

Pr=

1 Before a man’s foreskin is removed during the MC procedure, an injection is given at the base of the penis to prevent pain [TRUE]

100 (100–100)

99.4 (98–100)

99.6 (99–100)

0.4359

0.509

93.4 (90–97)

93.3 (92–95)

93.3 (92–95)

0.0054

0.941

2 It is possible to have pain, swelling, and bleeding after the MC procedure, but resting for 1–2 days after the surgery will help the wound heal [TRUE]a

92.8 (86–99)

95.6 (92–99)

94.7 (92–98)

0.7805

0.377

94.4 (91–98)

97.8 (97–99)

97.1 (96–98)

6.0953

0.014

3 There are no risks in MC surgery [FALSE]b

56.5 (45–69)

71.1 (64–78)

66.7 (61–73)

4.5824

0.032

87.3 (83–92)

86.8 (84–89)

86.8 (84–89)

0.0396

0.842

4 A man who is circumcised no longer needs to use condoms during sex to prevent him from becoming infected with HIV [FALSE]

89.9 (83–97)

95.6 (93–99)

93.9 (91–97)

2.7531

0.097

88.8 (84–93)

98.7 (98–100)

96.6 (96–98)

46.6747

0.000

5 All circumcised men are HIV negative [FALSE]

85.5 (77–94)

94.3 (91–98)

91.7 (88–95)

4.9140

0.027

84.8 (80–90)

90.3 (88–93)

89.2 (87–91)

5.0338

0.025

6 An HIV-negative man who is circumcised should continue to reduce his number of sexual partners to lower his chance of getting HIV [TRUE]

92.8 (86–99)

94.3 (91–98)

93.9 (91–97)

0.2100

0.647

92.4 (89–96)

96.6 (95–98)

95.7 (94–97)

6.6165

0.010

7 MC can help lower a man’s chances of getting penile cancer [TRUE]

92.8 (86–99)

83.7 (78–89)

86.4 (82–91)

3.3962

0.065

85.3 (80–90)

87.3 (85–90)

86.9 (85–89)

0.5609

0.454

8 A circumcised man who is HIV positive cannot pass HIV to his female partner [FALSE]

88.4 (81–96)

87.4 (82–93)

87.7 (83–92)

0.0433

0.835

78.7 (73–84)

92.1 (90–94)

89.3 (87–91)

29.2871

0.000

9 A man can start having sex after being circumcised when he feels better, even if it is sooner than 6 weeks after his MC surgery [FALSE]

92.8 (86–99)

88.7 (84–94)

89.9 (86–94)

0.8807

0.348

90.4 (86–95)

96.4 (95–98)

95.2 (94–97)

12.5475

0.000

10 MC increases a man’s chance of getting some STIs [FALSE]

85.5 (77–94)

87.4 (82–93)

86.8 (82–91)

0.1543

0.694

86.3 (81–91)

95.6 (94–97)

93.7 (92–95)

23.1110

0.000

Proportion passing test

89.9 (83–97)

88.1 (83–93)

88.6 (84–93)

0.1551

0.694

85.3 (80.3–90.3)

95.5 (94.0–97.0)

93.4 (91.8–95.0)

26.4573

0.000

Mean score

8.8 (8.4–9.2)

9.0 (8.8–9.2)

8.9 (8.7–9.1)

−1.0467

0.296

8.8 (8.6–9.0)

9.4 (9.3–9.4)

9.2 (9.2–9.3)

−6.2626

0.000

aAlternative wording in Swaziland: Follow-up visits at the clinic after MC surgery are to make sure the wound is healing properly [TRUE]

bAlternative wording in Swaziland: Risks of the male circumcision procedure include pain, swelling, or bleeding during or after surgery [TRUE]

In Swaziland, significantly fewer adolescents passed the test than adults (86 vs. 96 %, p < 0.001) and adolescents also had significantly lower mean scores (8.8 vs. 9.4, p < 0.001) than adults. In Zambia, although pass rates and mean scores were similar between adults and adolescents, 57 % of adolescents versus 71 % of adults (p < 0.05) responded correctly to the question about surgical risks (Table 2, Question 3) and 86 % of adolescents versus 94 % of adults (p < 0.05) responded correctly to the question, “A circumcised man who is HIV positive cannot pass HIV to his female partner” (Table 2, Question 8). There were no significant differences in pass rates or mean scores by survey language and, in Swaziland, there were no significant differences by clinic type, location, or affiliation (data not shown).

Separate multiple logistic regression models were fitted for each country, containing most of the same independent variables, but with some differences based on relevant factors in each country (Table 3). In Zambia, three variables were significantly associated with passing the test: participants who were able to read the entire sentence of the literacy test had six times the odds of passing compared to those who could not read some or all of the sentence (AOR 6.09; 95 % CI 1.52, 24.37); those who had completed at least upper secondary education had nearly five times the odds of passing than those who completed junior secondary education or less (AOR 4.70; 95 % CI 1.32, 16.67); and those counseled in the language in which they were most comfortable had nearly three times the odds of passing compared to those counseled in another language (AOR 2.84; 95 % CI 1.02, 7.94). In Swaziland, passing the literacy test was also a significant predictor of passing the comprehension test (AOR 2.92; 95 % CI 1.30–6.60) and adults had more than three times the odds of passing compared to adolescents (AOR 3.45; 95 % CI 1.35–8.77). Study participants from Swaziland who heard about MC from a friend compared to any other referral source had half the odds of passing (AOR 0.44; 95 % CI 0.24–0.80).
Table 3

Logistic regression: factors associated with passing the true/false comprehension test, by country

 

Zambia (n = 228)

Swaziland (n = 953)

UOR

AOR

95 % CI

p value

UOR

AOR

95 % CI

p value

Age group [adolescents (13–17)]

 Adult (18+)

1.02

0.51

0.13–1.97

0.329

3.67

3.45

1.35–8.77

0.009

Education [none]

 Some primary (grade 1–7)

n/a

n/a

n/a

n/a

1.19

0.60

0.13–2.82

0.514

 Secondary (form 1–3)

n/a

n/a

n/a

n/a

2.79

0.86

0.17–4.40

0.872

 High school (form 4–5)

n/a

n/a

n/a

n/a

7.78

2.05

0.37–11.32

0.410

 Higher than secondary

n/a

n/a

n/a

n/a

7.50

2.20

0.33–14.31

0.408

 Upper secondary or higher versus none through junior secondary

4.28

4.70

1.32–16.67

0.017

n/a

n/a

n/a

n/a

 Able to read entire sentence [no]/yes

7.72

6.09

1.52–24.37

0.011

4.69

2.92

1.30–6.60

0.010

 Passed numeracy test [no]

n/a

n/a

n/a

n/a

0.27

0.65

0.13–3.21

0.598

Marital status [single/never married]

 Currently married or living together

n/a

n/a

n/a

n/a

0.35

0.44

0.09–2.20

0.316

 Married

0.80

0.72

0.17–3.07

0.655

n/a

n/a

n/a

n/a

 Separated, divorced, widowed, other

0.48

0.28

0.02–3.56

0.326

1.81

1.12

0.43–2.93

0.816

Recruitment

 Heard about MC from friend [no]/yes

n/a

n/a

n/a

n/a

0.40

0.44

0.24–0.80

0.007

 Heard about MC from media [no]/yes

n/a

n/a

n/a

n/a

2.00

1.42

0.73–2.75

0.306

 Heard about MC from school [no]/yes

n/a

n/a

n/a

n/a

0.73

1.26

0.56–2.86

0.577

 Heard about MC from parent [no]/yes

n/a

n/a

n/a

n/a

0.32

0.65

0.27–1.58

0.344

 Recruited by IPC agent [no]/yes

n/a

n/a

n/a

n/a

0.95

1.15

0.60–2.21

0.679

 Ever received services from the clinic before MC [no]/yes

0.68

0.42

0.11–1.68

0.221

1.26

1.05

0.34–3.22

0.937

Clinic type

 Fixed versus [outreach]

1.49

2.48

0.62–9.96

0.200

0.88

2.53

0.92–6.95

0.071

 Urban versus [rural]

n/a

n/a

n/a

n/a

0.40

0.31

0.06–1.71

0.179

 High volume versus [low]

0.58

0.28

0.77–1.04

0.058

n/a

n/a

n/a

n/a

Counseling

 Received all counseling (group, 1-on-1, CT) [did not receive all counseling]

n/a

n/a

n/a

n/a

2.07

2.11

0.90–4.94

0.904

 Counseling included 1-on-1/CT [counseling included 1-on-1/CT plus group]

1.55

2.24

0.67–7.5

0.193

n/a

n/a

n/a

n/a

 Counseled in most comfortable language during group [no]/yes

n/a

n/a

n/a

n/a

1.75

1.69

0.76–3.69

0.198

 Counseled in most comfortable language, 1-on-1 [no]/yes

n/a

n/a

n/a

n/a

1.86

1.21

0.58–2.53

0.609

 Counseled in most comfortable language [no]/yes

2.88

2.84

1.02–7.94

0.047

n/a

n/a

n/a

n/a

Decision-makinga

 Decision comfort [not comfortable]/Very comfortable with decision

n/a

n/a

n/a

n/a

1.83

1.30

0.71–2.38

0.390

 Moderately comfortable with decision [very comfortable]

1.16

0.77

0.17–3.51

0.736

n/a

n/a

n/a

n/a

 Somewhat uncomfortable with decision [very comfortable]

1.50

0.76

0.14–4.13

0.751

n/a

n/a

n/a

n/a

 Very uncomfortable with decision [very comfortable]

2.03

1.21

0.12–12.75

0.873

n/a

n/a

n/a

n/a

 Felt could change mind? [no]/yes

2.97

1.89

0.59–6.04

0.281

1.54

0.89

0.49–1.64

0.712

n/a data not available/collected in specified country, UOR unadjusted odds ratio, AOR adjusted odds ratio, CT HIV counseling and testing

aIn Zambia, the reference category was “not comfortable” whereas in Swaziland, the reference category was “very comfortable.” Therefore, the categories and results for the two countries differ

Semi-Structured Interviews, One-Week Post-MC

A total of 94 SSIs were conducted 1 week after the MC procedure (n = 64, Zambia: 34 adults, 30 adolescents; n = 30, Swaziland: 16 adults, 14 adolescents). Two adolescent interviews from Zambia were excluded due to irregularities in the transcripts.

Background Demographics

As shown in Table 4, the majority of SSI respondents were single and under 25 years old. The majority of adults in Zambia had completed upper secondary education (86 %), compared to slightly more than half of the adults in Swaziland (56 %). In both countries, most adolescents were enrolled in secondary school during the study (data not shown). In Swaziland, most interviews were conducted in Siswati or a mix of Siswati and English with only 7 % conducted in English only. By contrast, the majority of interviews in Zambia were conducted in English (61 %), with less than half conducted in a mix of languages and 16 % conducted in Nyanja only.
Table 4

Characteristics of clients participating in semi-structured interviews, by country and by age group

Characteristica

Zambia

Swaziland

 

Adolescent (n = 28)

%

Adult (n = 34)

%

All (n = 62)

%

Adolescent (n = 14)

%

Adult (n = 16)

%

All (n = 30)

%

Mean age in years (range)

15.7 (13–17)

26.2 (18–43)

21.4 (13–43)

15.1 (13–17)

23.6 (18–43)

19.6 (13–43)

Single

97

59

78

100

81

90

Completed upper secondary education or higher

14

86

51

0

56

30

Language of interviewb

 English only

50

71

61

0

13

7

 English and Bemba

0

6

3

n/a

n/a

n/a

 English and Nyanja

21

21

21

n/a

n/a

n/a

 Nyanja only

32

3

16

n/a

n/a

n/a

 English and Siswati

n/a

n/a

n/a

36

44

40

 Siswati only

n/a

n/a

n/a

64

44

53

n/a Data not relevant in specified country

aDemographic data for semi-structured interview participants was collected by interviewers in the context of the discussions rather than in a structured survey; therefore, information available for the participants interviewed is not as detailed as for those who completed the quantitative comprehension test

bDue to rounding, percentages may add up to more than 100 %

Risks of MC Surgery

Among the 92 respondents, almost all said there were risks associated with the MC procedure. A few respondents described how the potential harm or discomfort resulting from the surgery or healing period factored into their decision-making process, but did not necessarily associate these potential adverse outcomes with the term “risk.” For example, a 27-year-old from Swaziland said he had forgotten the risks discussed during counseling, yet later in his interview he noted that bleeding or swelling would lead him to seek help. Some respondents reported not having been told about risks during counseling, but were aware of risks because they had learned about them from other sources such as friends and family. A few respondents described an active process of weighing the risks and benefits of MC, noting that risks were short-lived and worth dealing with, compared with long-term benefits. Only four respondents (all in Zambia) said they were unaware of any risks involved in the MC surgery.

The most common risk mentioned was pain, which was often anticipated prior to coming for MC surgery.

He said definitely pain was going to be there, yes, that’s basically one of the things [risks]…and then one of the negative things that he said was that…before you are operated on there’s going to be an injection…that can also inflict a bit of pain, just a bit of pain for them to start the operation properly.

30-year-old, Zambia

Adolescents in both countries seemed particularly concerned about the consequences of not caring for the post-surgical wound properly, referring to the potential for “rotting” (infection). A 15-year-old from Swaziland said, “I must keep the bandage dry because if it [gets] wet, the wound will be rotten” and a 13-year-old from Swaziland said, “What can I do if the wound will not heal or becomes rotten, such that I lose the whole penis?”

Some respondents seemed to conflate the risks of the MC surgical procedure with reduced risk of HIV post-MC. Several respondents said that a risk of MC was thinking one was now protected from getting HIV and would engage in behavior that increased risk of HIV (multiple partners, not using condoms). A 21-year-old from Swaziland said that the only risk the counselor had told him about was “the misconception that men usually have, that is, thinking they can no longer use condoms.” A 32-year-old from Zambia explained, “Not just because you have been circumcised it means you are 100 % [safe] from contracting diseases and you start having sex anyhow and say you can’t contract any disease, the risks are there for you to contract, so if you abstain continue doing that.”

Most sexually-active respondents (all adults, and some Zambian adolescents) seemed to understand that resuming sexual relations before the wound was completely healed would be risky; however, respondents emphasized injury or a lengthy healing process to be the major risks. Very few adults (1 in Zambia, 2 in Swaziland) clearly articulated that engaging in sexual intercourse within the six-week healing period could increase HIV risk for themselves and their partners.

Benefits of MC

Benefits of MC were mentioned spontaneously by most respondents, particularly in the context of discussing motivations for circumcision. Most participants listed several benefits, with prevention of HIV and STIs mentioned most frequently. In Zambia, however, some respondents did not view HIV prevention as a benefit because MC provides only partial protection. A 13-year-old said, “You will prevent yourself from contracting STIs, but you can still get AIDS if you still play around with girls.”

Often, respondents reported that they had come for MC for one benefit, but learned of others during counseling.

When I came here, I wanted to reduce my chances of getting HIV, but the counselor told me about other benefits…she told me that MC also reduces chances of getting STIs…I think the issue of STIs is more important because if you keep on getting STIs, that may lead into high risks of getting HIV.

23-year-old, Swaziland

Many respondents also listed hygiene as a benefit. Several respondents equated improved hygiene with reduced susceptibility to disease and staying healthy.

I knew that if I got circumcised I wouldn’t be contracting diseases and again I just felt that getting circumcised is a good thing…because there are times when certain diseases and lack of hygiene also causes diseases…so I just wanted to be clean.

15-year-old, Zambia

Many respondents in Zambia, but only a few in Swaziland, listed prevention of cervical cancer as a benefit. However, very few (3 in Zambia, 1 in Swaziland) mentioned that MC can reduce the risk of penile cancer.

Some respondents, primarily adults, included potential improvements in the sexual experience to be benefits of MC, and talked about prolonged erections, prevention of premature ejaculation, increased pleasure for themselves or their partners, and reduced pain during sexual intercourse. A 23-year-old from Swaziland said he had heard rumors in the community that “when you are circumcised, you perform better in bed” although he noted that the counselor had said that this had not been scientifically proven.

Partial Protection Against HIV and STIs

All of the adults in both countries said MC is not 100 % protective against HIV, the majority of whom correctly reported the estimated level of protection to be 60 %. However, some men may have been repeating the number they had heard, without actually understanding what “60 % reduced risk” meant. For example, two 21-year-olds from Zambia seemed to confuse whether 60 % of men were protected or at risk of getting HIV after MC. One explained that “the information they were telling us was that if you got circumcised you will be free from HIV and AIDS but not 100 % free; it will be about 40 % or so” and the other said “MC reduces HIV by…I don’t [know]—by 60 or 45 %.”

Many respondents, particularly in Swaziland, seemed to emphasize that removal of the foreskin, which they understood to facilitate HIV transmission, meant that MC would enable males to be “clean” and would eliminate places for diseases to “hide” in the body. A 23-year-old said a counselor had told him “some diseases hide in the foreskin and they penetrate through your body. If you are circumcised, those viruses won’t have a chance to get in you.” Similarly, a 13-year-old said “If you’re not circumcised, the foreskin acts as a hiding place for infections and the virus can easily get into you, but if you are circumcised, there is no place for infections to hide.” No one in either country articulated that HIV is transmitted during sex through semen, vaginal secretions, or blood, even after the foreskin has been removed.

Although nearly all respondents talked about MC being partially protective against HIV, there was confusion about the level of protection against other STIs. Many thought MC reduces the risk of other STIs (and cervical cancer) by 60 %, whereas others seemed to think that MC completely eliminates the risk of STIs and partners’ risk of cervical cancer, while only partially protecting against HIV. A 35-year-old from Zambia explained that “circumcision will only protect partners from cancer and other STIs, but HIV it will not protect you, you can still contract it.” In both countries, several adolescents seemed to think MC was 100 % protective against both HIV and other STIs. Even those who were fairly accurate about the level of protection against HIV afforded by MC, however, seemed to have difficulty applying the information to their own level of risk. Adolescents, in particular, talked about the need to continue safer sex practices, yet seemed to believe they were now at low (versus lower) risk of getting HIV post-MC.

They said it’s only 60 %, so you still have to wear a condom during sex, safe sex… I find that MC is there so in my future I wouldn’t have to worry much about getting HIV and AIDS and these sexually transmitted infections.

13-year-old, Zambia

Only a few adults (and no adolescents) seemed to have a clear understanding of partial protection and the risk of HIV infection post-MC.

Because when circumcised, it doesn’t mean you can’t get the virus… when they say 60 % it doesn’t mean 100 %. You can be among the 40 % who could get infected. Six out of ten, you can be the unlucky four. It’s a game of chance, it’s a game of dice, it’s a game of poker—you never know. You could win or lose. Chances are good that you are going to succeed, but chances are also that you may not…you are going to be unlucky. So, that is why it’s always good that you wear a condom.

29-year-old, Zambia

Impact of MC on Women

There was significant confusion regarding the lack of protective effect against HIV for female partners of circumcised men. Many respondents in both countries said that the impact of MC on HIV risk for female partners had not been discussed during counseling. The majority either thought MC had some protective effect against HIV for their partners, or did not know. Some equated the 60 % risk reduction to apply to their partners, as explained by a 35-year-old from Zambia who said, “Yes, I am told it protects [female] partners by 60 %.” A 21-year-old from Swaziland said, “If I am protected it means that she is automatically protected.”

Several respondents believed MC protects women, but only indirectly. For example, a 26-year-old from Swaziland said, “The woman will benefit in the sense that the partner has less chances of being infected. As a result, the woman will receive less infection.” A minority of participants (all adults) clearly understood that MC does not protect female partners against HIV. A 30-year-old from Zambia said, “For instance, if a circumcised man has HIV, a woman is not protected, but that 60 % only applies to circumcised men.” A 23-year-old from Zambia clearly conveyed that “if you are circumcised, you reduce the chances of causing cervical cancer in her when you have sex…There is no other protection to your partner if you are circumcised.”

HIV Risk-Reduction Strategies Post-MC

The majority of Zambians and all Swazi respondents (except one adolescent) acknowledged the need for continued HIV risk reduction practices post-MC. When asked how he would protect himself from HIV in the future, a 17-year-old from Zambia said, “I have to continue with my same practice—no changing of girls, condoms…using protection or staying away from sex, that’s the easiest one of all.” Most clients mentioned the importance of using condoms, even after circumcision. A 17-year-old from Swaziland said, “The counselor said it does not mean that you are safe when you are circumcised…the best thing to do is to abstain or else use a condom.” In a few cases, however, it was difficult to discern whether respondents thought condoms were only needed temporarily, during the immediate post-healing period, or that condoms should be used even after healing.

More than half of respondents discussed the importance of partner reduction after MC. Some noted that having multiple partners during healing would be especially risky. Some talked about MC as an opportunity to start a new life and that a negative HIV test served as further inspiration to get circumcised and be even safer in the future. A 40-year-old from Zambia said that after MC is “actually the time you reduce the number [of] partners, because it will be very embarrassing to say that we are circumcised, after that you have HIV….” More adults than adolescents emphasized monogamy as a post-MC risk-reduction strategy, whereas more adolescents, many of whom were not yet sexually active, focused on continued abstinence.

Discussion

This study was the first to evaluate the IC process for MC in Zambia and Swaziland in the context of VMCC programs going to scale. Although a majority of participants were able to pass the ten-question true/false comprehension test, several important concepts were not well understood by participants in both the quantitative and qualitative components of the study. Questions related to concrete information, such as anesthesia injections prior to surgery, safer sex practices, and wound care were more likely to be answered correctly than more nuanced concepts, such as “partial protection” or “risk.”

A key finding was the difference in comprehension between adults and adolescents. Even when controlling for education, in both countries, more adolescents than adults responded that there were no risks involved in the MC surgery, and in Swaziland, adolescents scored significantly lower than adults on the true/false test. It is possible that adolescents, many of whom made the decision to be circumcised within a few days of the procedure as a result of demand campaigns during school holidays, did not have as much information as adults, who may have been considering the procedure for longer. Adolescents may also have insufficient general life experience to comprehend complex information well enough to make an informed decision [38], which has implications for seeking the consent of younger males.

Despite overall high scores on the comprehension test, approximately one-third of participants (and nearly half of adolescents) in Zambia incorrectly responded that there are no risks involved in the MC surgery. Even in Swaziland, when the question was rewritten to be affirmative in light of emerging findings from Zambia (“Risks of the MC procedure include pain, swelling or bleeding during or after surgery”) 13 % of participants responded incorrectly.

Data from the SSIs may help to contextualize why the term “risk” was problematic for clients who took the comprehension test. The majority of SSI respondents could describe their fear of pain, bleeding, lengthy wound healing, or complications they had considered when choosing MC, yet they did not always associate these potential adverse outcomes with the term or concept of “risk.” An extensive body of research outlines the difficulties associated with communicating about “risk” during the IC process [3942]. Risk communication may be particularly difficult in the context of MC, because the term “risk” is used (in English and in local languages) to outline potential negative consequences of the MC surgical procedure while simultaneously describing the primary benefit of MC (reduced “risk” of HIV). Researchers and service providers should explore using different words or expressions to describe the MC surgical risks and reduced risk of HIV. It might be clearer if “risk” was used only in the context of describing potential adverse surgical outcomes and a different word, such as “chance” or “possibility” was used only to explain reduced probability of acquiring HIV.

Although most participants understood that MC is only partially protective against HIV, there was confusion about the specific level of protection MC affords against HIV, STIs, and cervical cancer. Many people thought MC completely eliminates the risk of STIs and cervical cancer, while only partially protecting against HIV. Others, however, thought the 60 % reduced risk applies to HIV, STIs, and cervical cancer. There is an extensive body of literature indicating that people, including those who are well educated, often have difficulty understanding numeric representations of risk and find it challenging to use this information to make health decisions [4046]. Research has shown that people often understand frequencies better than percentages (for example, “4 out of 10 circumcised men may nonetheless acquire HIV”) [40, 47, 48] and absolute numbers better than relative risk (for example, “In one year, 9 out of 1,000 circumcised men versus 21 out of 1,000 uncircumcised men will become infected with HIV”). Studies have also shown that pictorial or graphic representation of probabilities is more easily understood [43, 45, 49] and others have shown that providing less information is more effective than providing too much detail [50]. Research is needed to find the best ways of communicating about reduced risk of HIV post-MC; a simpler message about being at “lower” versus “low” risk of acquiring HIV/STIs might be more effective than providing percentages and relative risk information about HIV, and no specific information about other STIs.

Despite confusion about the exact level of protection afforded by MC, there was a high level of acknowledgment of the importance of continuing safer sex practices post-MC. More than 90 % of comprehension test participants answered correctly that circumcised males still need to use condoms and reduce their number of partners. In addition, nearly all SSI respondents discussed the need to continue safer sex practices—abstinence, fidelity, condom use—after MC, primarily because they understood MC to be only partially protective. All SSI participants understood the instruction to abstain from sex during the six-week healing period post-MC, however, very few men mentioned that not following this instruction could increase HIV risk. Most respondents explained the need to wait to ensure proper healing so as not to damage the wound and only rarely did respondents mention increased risk of HIV. It is critical that service providers emphasize more strongly that engaging in sexual relations within six weeks post-MC puts both clients and their partners at heightened risk of HIV.

Semi-structured interview participants indicated that pre-surgical counseling often excluded information about whether or not MC provides any protection for women against HIV. Therefore, it is not surprising that there was a general lack of awareness that MC does not directly reduce women’s HIV risk. More than 10 % of comprehension test participants thought that a circumcised man who is HIV positive cannot transmit HIV to his female partner, and only a minority of SSI participants seemed to know that MC protects women against HIV only indirectly. The lack of information provided may have led some study participants to incorrectly hypothesize that if MC reduces the risk of cervical cancer, then it must also mean circumcised men cannot transmit other STIs, including HIV, to their partners.

Limitations

This study had several limitations. First, study participants were not randomly selected but represented a convenience sample of MC clients agreeing to take the comprehension test or to be interviewed at participating sites. Therefore, it is possible that the data are biased in some unmeasured way. In addition, study participants (both quantitative and qualitative) were generally better educated than the populations, overall, in Lusaka or Swaziland [6, 7] limiting the generalizability of the findings; however, recent data from Zambia indicate that those coming for MC tend to have had more education than the general population [28], so this sample may be representative of males coming for MC. Regardless of how well the sample represents the broader population, if “risk” and “partial protection” were difficult concepts for clients with higher levels of education to comprehend, it is likely that those with lower levels of education will have even greater difficulty.

Conclusion

Voluntary medical MC represents an important strategy for reducing the impact of HIV and AIDS. To ensure autonomy and to maximize population level benefits, it is critical that clients choose MC with a full understanding of the risks and benefits of the procedure, the consequences of resuming sexual relations too soon, and the importance of continued HIV risk reduction post-MC. Research is needed to explore better ways to help clients distinguish between surgical risks and reduced risk of acquiring HIV post-MC, how to communicate the lack of protection against HIV for women, and to improve understanding among adolescents. A sound informed consent process is critical for maximizing benefits and minimizing potential risks for individuals and communities as VMMC programs continue scaling up.

Acknowledgments

We gratefully acknowledge the Zambian MOH, the University of Zambia Research Ethics Committee, and the Scientific and Ethics Committee of Swaziland for their review and approval of the study. The research team would like to acknowledge the collaborating institutions and their staff for their active involvement and support of this study; in particular, Steve Gesuale, Hayden Hawry, and Jessica Greene (Population Services International—Society for Family Health, Zambia); Marie Stopes International (MSI Zambia and MSI Swaziland); Family Life Association (FLAS) of Swaziland, and Jhpiego (Baltimore, MD). We are also grateful for the contributions of the study coordinators in Zambia (Kelvin Munjile) and Swaziland (Alfred Adams) and the data collection teams. We are especially grateful to the MC clients and to the parents and guardians of adolescent clients who generously gave their time to participate in the study. Significant support was provided by Population Council staff including Sharon Abbott, Maria Alevrontas, Nicole Haberland, Sherry Hutchinson, Virginia Kallianes, Barbara Mensch, Barbara Miller, and Naomi Rutenberg. This study was conducted under a subgrant from Population Services International (PSI) through the MC Partnership, which is sponsored by PSI with support from the Bill & Melinda Gates Foundation.

Copyright information

© Springer Science+Business Media New York 2013