Abstract
Objective:
The objective of this study was to assess whether HIV programming in southern Botswana could be leveraged to provide care for patients with noncommunicable diseases (NCDs).
Methods:
A retrospective analysis was performed to determine the spectrum and complexity of NCDs seen by HIV-focused outreach programming delivered between July 2011 and December 2013, to 9 facilities in southern Botswana. The association of HIV status and specific International Classification of Disease codes was examined using bivariate analysis.
Results:
Outreach HIV physicians recorded 926 outpatient consults involving 835 patients during the studied period. While 25% (n = 209) of patients seen were HIV infected, most patients were either HIV negative (49%, n = 410) or had an unknown HIV status (26%, n = 216). Noncommunicable disease referrals were as common at primary- and district-level facilities (90% [n = 459] versus 93% [n = 301]; P = .22).
Conclusion:
This study demonstrates how HIV programming in Botswana can be leveraged to improve access to specialist medical services for patients with NCDs.
Background
Despite a huge burden of infectious diseases such as tuberculosis and HIV/AIDS across sub-Saharan Africa, increasing incidence of chronic diseases such as diabetes mellitus and hypertension is placing strains on limited health resources. Recognized as having the second highest HIV/AIDS prevalence in the world, Botswana has established one of Africa’s most progressive programs for dealing with HIV. By 2010, Botswana achieved near universal access to antiretroviral (ARV) treatment, with more than 95% of eligible persons able to access therapy. 1 The speed and scope of Botswana’s ARV treatment scale-up have been attributed to a number of factors, not least the political will of the Botswana government. Key factors include significant donor investment, high-quality HIV-specific training, 2 and effective mentorship programs to facilities. 3 Although there are several published studies demonstrating the impact of donor funding in Botswana to improve HIV care and treatment outcomes, 2,4 data quantifying the impact of HIV donor funding on diseases other than HIV are sparse. 5
Since 2006, the Botswana-UPenn Partnership (BUP) supported the Botswana national HIV treatment program with US government support through Presidents Emergency Plan For AIDS Relief (PEPFAR). One aspect of BUP’s support has been its HIV outreach program, providing regular clinical and technical support at facilities in southern Botswana. This program was originally developed to support the management of patients with advanced HIV and complicated tuberculosis/HIV coinfection. However, at the request of local clinical staff, the support was extended to care for noncommunicable diseases (NCDs).
Objectives
The objective of this study was to assess the extent to which BUP’s HIV programming in southern Botswana was used to provide care for patients with NCDs and to assess the characteristics of the patients referred for specialist care at supported outreach facilities.
Methods
A retrospective analysis was performed to determine the spectrum and complexity of NCDs seen by HIV-focused outreach programming to 9 facilities in southern Botswana. Four supported sites were larger district-level sites and 5 sites were smaller, primary-level sites. All facilities had been identified by the Ministry of Health, based on the catchment area of HIV seroprevalence and lack of onsite medical specialists. At all facilities, the outreach medical specialists ran monthly clinics, seeing only patients referred by local clinicians.
Data Collection
Clinical and demographic data were abstracted from medical records for all outpatient encounters performed by outreach specialists at each site and included patients’ HIV status and International Classification of Disease diagnostic codes for all clinical encounters. For each encounter, primary and/or secondary diagnoses were recorded. For HIV-infected patients, HIV-related diagnoses were only recorded when the encounter pertained to HIV management.
Statistical Analysis
We entered data into an Access database (Microsoft Access; Microsoft Corporation, Redmond, Washington). Analysis employed Stata version 13.1, USA. 6 Comparisons between HIV-infected patients, HIV-uninfected patients, and those with an unknown HIV status were conducted using t tests for normal distributed continuous variables and chi-square tests for categorical variables. We report associations significant at the P < .01 level (2-tailed). If a patient was encountered more than once, we limited our analysis to data collected from the program’s most recent encounter with the patient.
Study Approval
The study was approved by Institutional Review Boards of the University of Botswana and the University of Pennsylvania as well as the Human Research and Development Committee of Botswana’s Ministry of Health.
Results
Between July 1, 2011, and December 31, 2013, the outreach physicians recorded 926 outpatient encounters involving 835 patients. Nearly half of all outreach encounters were with HIV-uninfected patients (49%, n = 410), 25% (n = 209) of those seen were HIV infected, and the remainder had an unknown HIV status (n = 216; Table 1). Most patients seen were women (59%, n = 496) and the mean age was 46.5 years. Patients with HIV were significantly younger than those with an unknown or an HIV-negative status (41.2 years versus 51.3 years versus 46.7 years, respectively; P < .001). Individuals with HIV were also far more likely to present with 2 or more medical problems than those with no HIV infection or those with an unknown HIV status (43% [n = 89] versus 14% [n = 588] versus 25% [n = 55], respectively; P < .001).
Characteristics of Outpatients Seen on Outreach.
Abbreviations: SE, standard error; CNS, central nervous system.
a Total exceeds 100% as some patients had multiple diagnoses.
More patients were encountered at primary hospitals than at district general hospitals (61% [n = 509] versus 39% [n = 326]). The proportion of patients presenting with multiple morbidities was higher at district compared to primary hospitals (32% [n = 103] versus 19% [n = 99], respectively; P < .001). There was no difference in the prevalence rate of NCDs encountered at district- and primary-level facilities (P = .22).
HIV-Infected Patients
Most HIV-infected patients were on antiretroviral therapy (ART) at the time of the encounter (81%, n = 168). Most were also referred for the management of concurrent NCDs (79%, n = 165) rather than for management of their HIV per se. There was no difference in referrals involving NCDs between patients on ART and ART-naive patients (77% [n = 130] versus 85% [n = 33], respectively; P = .32).
Noncommunicable Diseases
Regardless of the HIV status, the vast majority of patients referred presented with a diagnosis of one or more NCDs (91%, n = 761). Uncontrolled hypertension was the most common referral diagnosis among patients (13%, n = 140), followed by ischemic heart disease (6%, n = 53) and diabetes mellitus (DM; 6%, n = 46). Patients were also frequently referred with congestive cardiac failure (CCF; 5%, n = 42), migraines (3%, n = 27), and malignancies (2.7%, n = 23).
Discussion
This analysis supports other data demonstrating that there is a significant burden of NCDs both among people with HIV and in the general population in southern Africa. 7,8 Notably, four-fifths of patients referred for management of their non-HIV medical problems were already on ART. Plausibly, the reason that so many patients referred to specialists were presenting with complex NCDs is due to the fact that they are living longer and developing metabolic complications of highly active antiretroviral therapy (HAART). This seems consistent with the evolution of HIV/AIDS programs in high-income countries and in other African settings, where the prevalence of non-AIDS-defining illnesses, such as diabetes and dyslipidemia, has risen commensurate with access to HAART. 9 Our analysis also suggests that there is a significant burden of NCDs in the general population in Botswana. This is consistent with local 10 and regional epidemiologic studies reporting an increasing burden of NCDs in southern Africa. 11,12
We speculate that outreach physicians were referred many patients with NCDs because local referring clinicians felt uncomfortable managing patients with complex chronic diseases. While Botswana has more doctors per capita than many other countries in southern Africa, 13 many may lack the training and confidence to manage complex NCD. A cogent response to the emerging NCD epidemic will need to incorporate NCD training into current HIV capacity building initiatives and enhance integration of NCD and HIV clinical services at facility level. An integrated approach to NCD and HIV care will also need considerable investments in pharmacy supply-chain management and the development of point-of-care diagnostics. However, it may have considerable benefits, reducing transport barriers and staffing costs. 14
Limitations
This study is limited by referral bias, since we only report on patients who were referred to and evaluated by outreach physicians. We excluded inpatients also seen by HIV physicians on outreach, who had more severe manifestations of disease. We acknowledge that the study is limited by the type of data collected, since patient encounters were recorded for programmatic not research purposes. Given the cross-sectional nature of the data, we were unable to demonstrate any trend in NCD incidence over time. Nevertheless, the analysis provides valuable insights into how NCDs pose significant challenges to existing primary care infrastructure in Botswana.
Summary
Botswana faces a double burden of infectious and noninfectious diseases. This study demonstrates how HIV programming can be leveraged to improve access to specialist medical services for patients with NCDs. Further research is necessary to identify optimum, cost-effective ways to integrate services for NCDs and HIV in high HIV prevalent settings in sub-Saharan Africa.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This publication was made possible through support provided by Presidents Emergency Plan For AIDS Relief (PS001949-05) and through core services and support from the Penn Center for AIDS Research (CFAR), an NIH-funded program (P30 AI 045008).