Screening for diabetes and hypertension in a rural low income setting in western Kenya utilizing home-based and community-based strategies

Background: The burdens of hypertension and diabetes are increasing in low- and middle-income countries (LMICs). It is important to identify patients with these conditions early in the disease process. The goal of this study, therefore, is to compare community- versus home-based screening for hypertension and diabetes in Kenya.Methods: This was a feasibility study conducted by the Academic Model Providing Access to Healthcare (AMPATH) program in Webuye, a town in western Kenya. Home-based (door-to-door) screening occurred in March 2010 and community-based screening in November 2011. HIV counselors were trained to screen for diabetes and hypertension in the home-based screening with local district hospital based staff conducting the community-based screening. Participants >18 years old qualified for screening in both groups. Counselors referred all participants with a systolic blood pressure (SBP) ≥160 mmHg and/or a random blood glucose ≥7 mmol/L (126 mg/dL) to a local clinic for follow-up. Differences in likelihood of screening positive between the two strategies were compared using Fischer’s Exact Test. Logistic regression models were used to identify factors associated with the likelihood of following-up after a positive screening.Results: There were 236 participants in home-based screening: 13 (6%) had a SBP ≥160 mmHg, and 54 (23%) had a random glucose ≥ 7 mmol/L. There were 346 participants in community-based screening: 35 (10%) had a SBP ≥160 mmHg, and 27 (8%) had a random glucose ≥ 7 mmol/L. Participants in community-based screening were twice as likely to screen positive for hypertension compared to home-based screening (OR=1.93, P=0.06). In contrast, participants were 3.5 times more likely to screen positive for a random blood glucose ≥7 mmol/L with home-based screening (OR=3.51, P<0.01). Rates for following-up at the clinic after a positive screen were low for both groups with 31% of patients with an elevated SBP returning for confirmation in both the community-based and home-based group (P=1.0). Follow-up after a random glucose was also low with 23% returning in the home-based group and 22% in the community-based group (P=1.0).Conclusion: Community- or home-based screening for diabetes and hypertension in LMICs is feasible. Due to low rates of follow-up, screening efforts in rural settings should focus on linking cases to care. © 2013 Pastakia et al.; licensee BioMed Central Ltd.

This pilot study is an initiative of the Academic Model Providing Access to Healthcare (AMPATH) program located in western Kenya, which is a partnership between Moi University, Moi Teaching and Referral Hospital and a consortium of North American universities led by the Indiana University School of Medicine. The history, organizational structure, and health programs of AMPATH have been described elsewhere [9-11]. AMPATH has the stated goal to build upon an academic foundation that supports all 3 academic missions of service (through healthcare), teaching, and research. AMPATH delivers care, provides education, and performs research in networks of urban and rural Ministry of Health hospitals, health centers, and dispensaries in western Kenya. AMPATH has also developed considerable infrastructure for providing home-based and community-based counseling and testing for HIV and has pioneered several innovative strategies for integrating intensified case finding into its comprehensive prevention strategy [9,12]. AMPATH delivers a comprehensive, community-based care program that initially focused on patients infected with HIV but has since expanded to provide maternal and child health services and chronic disease management (specifically diabetes and hypertension) to a catchment population of over 2 million persons [2]. This specific feasibility testing of screening for hypertension and diabetes was carried out in the rural town of Webuye which is located within the AMPATH catchment area. Webuye town (population 19,600 in 2011) was selected to carry out the pilot because home-based counseling and testing (HBCT) for HIV was underway at the time of the study [9,12]. There also exists accessible infrastructure for long-term care of diabetes and hypertension at Webuye District Hospital to manage all the positively screened patients [13]. This project was approved by the Institutional Research and Ethics Committee based at Moi University School of Medicine. The home-based screening pilot was carried out in March 2010 alongside the ongoing HBCT for HIV. Five home-based counselors (trained in HIV testing and counseling) with at least a high school education who were contemporaneously providing home-based HIV screening underwent a one-day training session on hypertension and diabetes. This training highlighted the epidemiology, pathophysiology, risk factors, diagnosis/measurement and screening methods for hypertension and diabetes. Counselors were also instructed on the overall goals of treatment and prevention options. Lastly, in practical sessions, the counselors were taught about the appropriate methodology for performing finger pricks, proper use of glucose testing strips and meters (Abbott Optimum Xceed), and proper use of an automatic sphygmomanometer (OMRON HEM-712c) with a medium sized cuff (22 to 32 cm). Each counselor was instructed, observed and required to demonstrate proper diabetes and hypertension screening techniques. After completion of the training session, investigators (SP, SMA, COA, GSB) accompanied counselors into at least one household to ensure that they were able to provide appropriate counseling and to measure blood pressure and blood sugar using the proper technique. All individuals 18 years of age and above who verbally consented to have their blood pressure and sugar measured were eligible. Counselors traveled door-to-door to offer screening services using a standardized protocol. After receiving verbal consent from the participants, counselors counseled and tested participants for HIV. Participants were also counseled on hypertension and diabetes screening and subsequently tested for random blood sugar (RBS). Blood pressure was measured at the end of the home visit after the counselor had tested for blood sugar and HIV. Blood pressure was measured only once unless an error message was recorded. In the case of an error message, most participants had their blood pressure measured again. If an error message was persistently observed, participants were instructed to follow up at the specified confirmation site at Webuye District Hospital. The community-based screening strategy was carried out in November 2011 in conjunction with Webuye District Hospital staff. Community mobilizers were used to sensitize the community to the availability of a 2-day long diabetes and hypertension screening program in the upcoming week to commemorate World Diabetes Day. The availability of this free screening was advertised through standard modalities of sensitization including discussion at church, via community chiefs meetings (locally referred to as Barazas), and by word of mouth via community leaders. Any person 18 years of age and above voluntarily visiting the screening booth received a free blood pressure check, free blood sugar test, had their height and weight recorded, and BMI calculated. The screening booth was situated in the center of town and was marked with signage and staff who actively advertised for the screening. The screening booth was easily visible and accessible to any participant interested in a free screening. Nurses and clinical staff with experience in the management of chronic diseases at the Webuye District Hospital chronic disease clinics were responsible for performing all elements of the screening program. The same screening and referral protocol was used for both the home-based and community-based screening strategies. Age, sex, medical record number (if available), contact information, HIV screening result (only performed in home-based screening), blood pressure screening result, and blood sugar screening result were recorded manually for all participants by the counselors on a standardized data collection sheet. In the community-based screening event, screening staff also recorded the height and weight on the same data collection sheet. A cut-off systolic blood pressure (SBP) of ≥ 160 mmHg was used in order to triage screened participants for referral to the local clinic for diagnostic testing. This cut-off was based on the limitation that only one blood pressure reading was obtained during screening and to avoid unnecessary referral due to one isolated high reading and regression towards the mean. In addition, previous studies in sub-Saharan Africa have used a higher screening threshold to avoid excess referral in settings with resource constraints to maximize the use of resources [14-18]. A diastolic blood pressure (DBP) cut-off was not used to screen participants [19]. Participants who met referral criteria were referred to the local clinic located at Webuye District Hospital for follow-up blood pressure measurements. At that follow-up visit, two blood pressure measurements were taken during the same visit. The average of the two blood pressures taken at the clinic was calculated. Hypertension was diagnosed based on the Joint National Committee VII criteria for systolic or diastolic blood pressure (SBP ≥ 140 or DBP ≥ 90 mmHg) [20]. A RBS cut-off of 7.0 mmol/L (126 mg/dL) was used as the threshold for referring participants for confirmatory testing in clinic. With the unpredictable nature of the timing of the screening, it is possible that some participants would have been fasting at the time of screening. Therefore, the recommended RBS cutoff of 11.1 mmol/L (200 mg/dL) for diagnosing diabetes would not have appropriately referred participants who arrived to the screening with a fasting blood sugar result ≥ 7.0 mmol/L. Because of this dynamic, the threshold for referral for all participants was set at 7.0 mmol/L to ensure that any patient who might meet either of these diagnostic criteria would be offered confirmatory testing. Patients meeting this referral threshold were then instructed to fast before coming for confirmatory testing at the hospital based outpatient diabetes clinic on a subsequent day. Patients with a fasting blood sugar above 7.0 mmol/L in the clinic were confirmed to have diabetes. Participants with a positive screen for hypertension (systolic blood pressure ≥ 160 mmHg) or diabetes (RBS ≥ 7 mmol/L) were provided an information sheet and referral card to follow up in clinic at the Webuye District Hospital. The participants who returned to the hospital based clinic were then provided with a free fasting blood sugar testing and/or 2 separate resting blood pressure readings to confirm the relevant diagnosis. Participants found to have SBP 140–159 mmHg during the initial screening were provided education on appropriate lifestyle modifications and dietary strategies, such as salt reduction. They were also instructed to obtain a follow-up blood pressure reading within 6 months at any local blood pressure testing facility. Participants with impaired fasting glucose (5.6 – 6.9 mmol/L or 100 – 125 mg/dL) on the clinic based confirmatory fasting blood glucose testing were also instructed to engage in lifestyle modifications and perform an annual fasting blood sugar at the nearest available facility. All participants receiving a confirmed diagnosis of diabetes or hypertension were instructed to engage in lifestyle modifications and were registered into the appropriate chronic disease clinic based on their diagnosis. Participants who did not follow up after having a positive screen for diabetes or hypertension in the initial screening were contacted via phone and provided additional directions and encouragement to visit Webuye District Hospital for confirmation. Participants were called on at least two separate occasions at both the primary and alternate phone number provided during the respective screening events. All participants with complete data recorded in the screening register were included in this analysis. Descriptive analyses were used to characterize the demographic characteristics of the findings of the two different screening strategies. To achieve the primary objective of comparing the feasibility of both strategies, descriptive analyses were used to assess the percentage of participants with an initial positive screen for diabetes and elevated SBP in both strategies. In addition, comparative statistical analyses were performed to identify statistically significant differences between the two strategies for diabetes and hypertension screening. The Fischer’s Exact Test was utilized to compare the difference in the likelihood of screening positive for diabetes or hypertension in the home-based versus community-based screening strategy. Additionally, the likelihood of following up after a positive screening test was compared between the two screening strategies and odds ratios (OR) for a positive screening were calculated. Exact logistic regression was performed to determine the characteristics associated with a positive screening for either diabetes or hypertension. Linear regression was utilized to identify the relationship of relevant covariates with blood pressure and blood sugar. In order to demonstrate the changes that might occur by setting more aggressive referral thresholds, additional analyses were completed to illustrate the additional numbers of patients that might be referred using different screening criteria. All analyses were completed using STATA® (College Station, Texas, Version 8).

N/A

• Dima Health