Frequency of visits to health facilities and HIV services offered to men, Malawi

Abstract Objective To determine how often men in Malawi attend health facilities and if testing for human immunodeficiency virus (HIV) is offered during facility visits. Methods We conducted a cross-sectional, community-representative survey of men (15–64 years) from 36 villages in Malawi. We excluded men who ever tested HIV-positive. Primary outcomes were: health facility visits in the past 12 months (for their own health (client visit) or to support the health services of others (guardian visit)); being offered HIV testing during facility visits; and being tested that same day. We disaggregated all results by HIV testing history: tested ≤ 12 months ago, or in need of testing (never tested or tested > 12 months before). Findings We included 1116 men in the analysis. Mean age was 34 years (standard deviation: 13.2) and 55% (617/1116) of men needed HIV testing. Regarding facility visits, 82% (920/1116) of all men and 70% (429/617) of men in need of testing made at least one facility visit in the past 12 months. Men made a total of 1973 visits (mean two visits): 39% (765/1973) were as guardians and 84% (1657/1973) were to outpatient departments. Among men needing HIV testing, only 7% (30/429) were offered testing during any visit. The most common reason for not testing was not being offered services (37%; 179/487). Conclusion Men in Malawi attend health facilities regularly, but few of those in need of HIV testing are offered testing services. Health screening services should capitalize on men’s routine visits to outpatient departments as clients and guardians.


Introduction
Men throughout sub-Saharan Africa have a lower life expectancy than women. 1 One contributing factor is men's underrepresentation in early disease detection, for example, timely diagnosis of life-threatening conditions such as tuberculosis, human immunodeficiency virus (HIV) infection, and hypertension. [2][3][4][5] Low uptake of routine screening services is concerning because delayed diagnosis can result in advanced stages of illness and increased risk of onward HIV or tuberculosis transmission. Effective strategies are needed to reach men with routine screening services.
Facility-based interventions for routine screening services can increase men's engagement in care and may be more scalable and sustainable than community-based approaches, particularly in resource-constrained settings where human resources are limited. 6,7 Yet, most interventions for men prioritize community-based approaches, [8][9][10] in part due to assumptions that men do not attend health facilities and intentionally avoid health facilities altogether. 11,12 A common belief is that community services are required to find so-called missing men who otherwise cannot be reached. 10,13 Little is known, however, about how often men actually attend health facilities, either as clients or in support of the health care of others (i.e. as guardians). Facility-based strategies during routine facility visits may be a feasible way to reach most men if men frequent health facilities.
While men in sub-Saharan Africa are less likely than women to access testing and treatment services for HIV and tuberculosis, 14,15 these services likely comprise only a minority of men's facility visits. Men may attend outpatient departments for acute care, or attend a variety of other departments as guardians to aid others seeking care, such as spouses or children. A study in Malawi found that 22% (90/401) of rural men had accessed health services for their own health in the previous 2 months. 16 However, the proportion of men attending health facilities over time is still unclear because findings were limited to visits for men's own health care (not as guardians) and only visits within the previous 2 months.
It is also unclear if men who attend health facilities for reasons other than HIV and tuberculosis services are offered screening services (such as HIV, tuberculosis or hypertension screening) during their facility visits. Research suggests that screening coverage during routine care may be low for a range of diseases, 17,18 especially among individuals attending outpatient departments that are often busy and overburdened. Low screening coverage could indicate missed opportunities to engage men already attending health facilities. Examining the frequency with which men attend health facilities, and whether screening services are offered when they do attend, is important to understand the potential role of facility-based interventions for men.
We used HIV testing as a case study to explore the potential role of facility-based screening strategies for men. We focused on HIV testing because Malawian national guidelines indicate that sexually active men should be tested for HIV every 12 months, regardless of symptoms or age. 19 HIV testing is also decentralized and widely available in nearly every

Research
Men's attendance at health facilities, Malawi Kathryn Dovel et al.
health facility in Malawi. We aimed to determine the frequency with which men in Malawi attend health facilities and the HIV services offered to them when they do attend.

Setting
In Malawi, an estimated 11% of men living with HIV are unaware of their HIV status compared with 6% of women. 20 According to national guidelines, facility-based testing should be offered through provider-initiated testing and counselling, whereby providers offer HIV testing services to clients in need of testing (defined as those who have never tested positive and did not test <12 months ago). 19 Implementation of provider-initiated testing and counselling varies widely by facility department, with nearly 100% coverage in antenatal clinics 21 and less than 15% testing coverage in outpatient departments. 22,23

Design and participants
We conducted a cross-sectional, community-representative survey of men in 36 villages in the central and southern regions of Malawi. We used a multistage sampling design to select study villages and potential respondents. We purposively selected two districts in central and southern regions (Lilongwe and Chikwawa) and three mid-size health facilities per district (six facilities in total). We then randomly selected six villages within each facility's catchment area by using a computer-generated sequence of random numbers (36 villages in total), and used household census listings to randomly select men within each village. Villages were a mean of 5.14 km (standard deviation, SD: 3.46 km) from facilities, had a mean of 72.8 (SD: 27.7) households per village, and the main occupations were farming, fishing and informal employment (further details available in the data repository). 24 Eligibility criteria for inclusion in the survey were: (i) age 15-64 years; (ii) current resident of the participating village; and (iii) spent > 15 nights in the village in the previous 30 days. We excluded men who did not meet eligibility criteria from the final household listing before randomization. For this paper, we also excluded men who self-reported ever testing HIV-positive. We stratified the random selection of men by village (about 45 men per village, although some villages had fewer than 45 men due to small village size) and age category: young men (15-24 years), middle-aged men (25-39 years) and older men (≥ 40 years). We used computerized random number generation to select strata.

Data collection and measurement
Male research assistants recruited randomly selected individuals with the assistance of community health workers and village chiefs. We categorized individuals as not found after three failed tracing attempts. We conducted surveys wherever convenient for respondents, which was usually at the respondent's home, the village chief 's residence (a typical gathering place in villages) or the respondent's place of work.
We collected the following data: (i) demographic data, such as age, marital status and education; (ii) selfreported HIV testing history, defined as tested recently (testing HIV-negative in the past 12 months) or in need of testing (never tested positive or tested > 12 months ago), as per national guidelines; 25 (iii) number of health facility visits made in the previous 24 months; (iv) services received during the last four health facility visits, including who received the services (i.e. client or guardian visit type); and (v) if HIV testing was offered and received during each facility visit. We defined client visits as any visit where the primary service received was for the respondent's own health. We defined guardian visits as any visit where the primary service was received by another person and the respondent attended the facility to support that person's use of health services -we did not categorize providing transportation and immediately leaving the facility as a guardian visit. We defined being offered HIV testing as being told about HIV testing by a health-care worker at the facility on the same day as a facility visit, and actual HIV testing as completing an HIV test the same day as a facility visit. We conducted surveys in the local language (Chichewa) and they lasted about 55 minutes on average.

Analysis and sample size
Our primary outcome was health facility attendance in the past 12 months -we excluded any health visits made more than 12 months before the survey. Secondary outcomes included being offered HIV testing during facility visits in the past 12 months and completing an HIV test at any facility visit in the same period. Our study included 1116 men (HIV-negative or with an unknown HIV status), which meant that it had had over 80% power to estimate population-level frequency of health-facility visits within the past 12 months, assuming 5% precision, 0.05 level of confidence and about 10 000 men and male adolescents in each facility catchment area.
We used descriptive statistics to examine how often men attend health facilities and the reason for facility visits (client or guardian), disaggregated by history of HIV testing (tested in the past 12 months or in need of testing). We conducted sensitivity analyses using assigned weights to account for variation in village size. The results from sensitivity analyses did not change findings and are available in the data repository. 24 No data were missing.

Ethical considerations
The National Health Sciences Review Committee of Malawi (number 2338) and the University of California Los Angeles Institutional Review Board (number 20-001606) approved study activities. All eligible individuals completed a written informed consent form immediately following screening procedures. For individuals between 15 and 18 years of age, guardians provided written consents.

Results
We recruited 1473 men between 15 August and 18 October 2019. Of these men, we screened 1293 (88%) men for eligibility -we were unable to screen 180 (12%) men ( Fig. 1). Of the 1293 men screened, 1117 (86%) were eligible for inclusion in the study. One man declined to participate in the study and thus we included 1116 men in the analysis (Fig. 1). Table 1 shows the demographic data of the respondents according to history of HIV testing. The mean age of men was 34 years (SD: 13.2), 84% (941/1116) felt healthy and 82% (910/1116) were married or living with a partner. Most men (69%; 767/1116) had worked in the previous 30 days and 27% (299/1116) had spent three nights or more away from their home in the same period.
Participants commonly reported risky sexual behaviour, with 31% (345/1116) reporting having two sexual partners or more in the past 12 months and 18% (198/1116) reporting not using a condom with at least one nonmarried or live-in partner in the same period. At the time of the study, 55% (617/1116) of the men needed HIV testing (defined as never tested or tested > 12 months ago) -26% (294/1116) had never been tested.
Of the total number of men included in the study, 82% (920/1116) had made at least one facility visit as either a client or guardian in the past 12 months, and 70% (429/617) of men in need of HIV testing had made at least one facility visit (Fig. 2). Secondary analyses show a similar frequency of facility visits among men who had never been tested (data repository). 24 We found 10% (109/1116) of all men and 11% (68/617) of men in need of HIV testing only attended facilities as a guardian (i.e. made no visits as a client).
A total of 1973 health facility visits were made by 920 men, with a mean two visits (SD 1-3) per participant, in the past 12 months (   Men's attendance at health facilities, Malawi Kathryn Dovel et al. reproductive health services or preventive services for children younger than 5 years. Among guardian visits, 42% (318/765) of men were accompanying children, 26% (199/765) were accompanying friends and 24% (182/765) were accompanying their sexual partner. Men in need of testing made significantly fewer facility visits, were more likely to attend a facility as a guardian (not a client), and were more likely to attend the outpatient department than men who had recently been tested for HIV (P < 0.001 for all visit characteristics). Men in need of testing were also significantly less likely than recently tested men to accompany their sexual partners during guardian visits (P < 0.001) -those in need of testing primarily accompanied children, friends or other relatives. Table 3 shows HIV testing services offered and used among men who attended a health facility. Among all men who attended a facility in the past 12 months, 25% (233/920) were offered provider-initiated HIV testing and counselling during at least one facility visit and 48% (441/920) were tested for HIV. A higher proportion were tested than offered testing because some men only attended facilities for HIV testing services or actively sought out HIV testing during their facility visit (without being prompted by a health-care worker). When excluding men who only attended facilities for HIV testing services (defined as only attended the HIV testing and counselling department, 143 men), 38% (298/777) of men were tested for HIV during a facility visit outside an HIV testing and counselling department. Fig. 3 depicts missed opportunities for reaching men with HIV testing. Among all men, the biggest gap in the cascade for facility-based testing was being offered HIV testing. While 82% (920/1116) of men had made at least one facility visit in the past 12 months, only 25% (233/920) were offered providerinitiated HIV testing and counselling (Fig. 3). Of those offered these services, 198 (85%) accepted them and were tested for HIV that same day. Among men in need of testing, 70% (429/617) had made at least one facility visit in the past 12 months, but only 30 (7%) were offered provider-initiated HIV testing and counselling, none of whom accepted testing.
The most commonly reported reasons for not testing during their most recent facility visit were: not offered testing (37%; 179/484); perceived low risk of infection (23%; 113/484); and not ready to test (17%; 81/484). Reasons reported were similar for client and guardian visits (data repository). 24 Among men making guardian visits, only 13% (35/268) were not willing to test because of lack of privacy because they were accompanying someone else.

Discussion
Few studies have examined men's general health-seeking behaviour in sub-Saharan Africa, or the role of men as guardians. Using data from a community-representative survey of 36 villages in Malawi, we show that more than three quarters of men with an HIVnegative or unknown HIV status had attended a health facility in the past 12 months. Over 80% of all visits to health facilities were to an outpatient department for acute, curative services; 39% of all visits were as a guardian to support the health care of others. Among men in need of HIV testing, 70% had attended a health facility in the past 12 months. Over 90% of their facility visits were to an outpatient department and 48% of visits were as guardians.
Our findings challenge the common belief that men do not attend health facilities, which has important implications for programmes targeting men. If facility-based services can be scaled up, most men may be reached during routine outpatient department visits. Targeted community-based strategies could then focus on the small proportion of high-risk men who do not regularly attend facilities, which could improve the efficiency and overall sustainability of HIV and other health screening programmes. However, facility-based services should capitalize on every visit made by men. Men in our study made a median of two facility visits in the past 12 months, far fewer than other studies have documented for women of similar ages. 26,27 Therefore, there may be relatively fewer opportunities to reach men at health facilities as compared with women, and every opportunity should be taken to engage men.
We found extensive missed opportunities to reach men who were already engaged with the health system. Among men in need of HIV testing who had attended a health facility, only 7% were offered HIV testing. Other research suggests that similar missed opportunities also exist for other routine screening services. 17,18 Given the multiple barriers to testing (such as long wait times, lack of privacy and unfriendly staff), 27-29 men may benefit from being offered testing during every facility visit, including at outpatient departments and during guardian visits. Being offered testing may substantially increase overall testing coverage. Of all the men in our sample, 85% of those offered testing ac- cepted, which is similar to other studies in the region. 30 We found that men primarily attended outpatient departments (for both client and guardian visits). This finding was especially true for men in need of testing where 93% of those who visited a health facility had visited an outpatient department. This result is in line with other research that find that few routine, universal health services are recommended for men who are not ill. 31 In Malawi, the only recommended preventive service for men younger than 45 years is annual HIV testing. 27 Outpatient departments should be a key place for engaging men in a range of health services and preventive care interventions.
A large proportion of men's facility visits were as guardians, where men supported the health care of others. Deliberate efforts to offer screening services to male guardians should be prioritized. Our findings differ from other studies that report that men are not involved in the health care of their families, 32 although recent literature recognizes that traditional gender roles for caregiving are changing. [33][34][35] Additional research is needed to understand men as guardians and how guardian visits can be best used as an entry point for men's own health.
How can interventions best capitalize on men's frequent facility visits? Optout services are key to improved uptake of a range of health services. [36][37][38] Screening services could be offered in outpatient department waiting areas while men wait for acute care, thus taking advantage of time already spent at the facility. Self-testing in outpatient department settings can also improve efficiency and minimize the human resources required to administer tests. 39,40 HIV self-testing at health facilities has been shown to reduce staff time required for testing, increase testing coverage and is acceptable among men. 22,41 Investment in adequate infrastructure and staffing may also be needed for longer-term solutions for screening efforts in outpatient departments.
Our study has several limitations. First, we used HIV testing as a case study to understand the potential reach of facility-based screening services among men. Missed opportunities for facility-based services may be higher for less prioritized health concerns, such as tuberculosis and hypertension. Second, our findings may not be generalizable outside Malawi as health-care-seeking behaviour may differ in countries with higher rates of formal employment and health insurance coverage, such as South Africa. Third, survey data rely on self-reporting and may be susceptible to social desirability bias if men believe they should engage in health services or HIV testing specifically. Fourth, we may have underestimated the proportion of men in need of HIV testing because we did not account for risk factors that increase the recommended frequency of testing -such as known HIV exposure or seeking services for sexually transmitted infections. Finally, our sampling frame did not account for variation in village size or the number of men within each age category in the general population. Sensitivity analyses assigned weights to adjust for these potential biases and found no differences in study results.