Capacity of Ugandan public sector health facilities to prevent and control non-communicable diseases: an assessment based upon WHO-PEN standards

Rapid population growth, urbanization, a nutritional transition, and indoor and outdoor air pollution are among the factors contributing to the growing burden of the major global non-communicable diseases (NCDs) in Uganda [1]. Based on a recent nationwide representative survey, the prevalence of hypertension, at 24.3%, approaches that in the United States [2]. The prevalence of diabetes mellitus is relatively low in urban (2.7%) and in rural (1%) areas, although it is expected to rise significantly in the near future [3]. Chronic obstructive pulmonary disease (COPD) in rural Uganda, where the prevalence is 16.2%, appears to be more common than in high-income countries [4]. The epidemiology of cancers is shifting from a predominance of HIV-associated malignancies to those more typically associated with “Western lifestyles” [5].

Despite this burden of risk factors and associated disease, as well as extensive experience in addressing chronic HIV infection, the Ugandan healthcare system remains ill-prepared to address NCDs [6]. Only a few prior studies have evaluated the capacity of public sector health facilties to provide NCD services. The Service Availability and Readiness Assessment (SARA) is a World Health Organization (WHO) -endorsed cross-sectional survey that provides insight into overall health facility readiness, though is limited in its acquisition of data specific to NCDs. To date, three SARA surveys have been conducted in Uganda: 2012, 2013, and 2014 with the latter being restricted to a sample of higher level facilties and conducted after the present study was completed [7]. The 2013 Uganda SARA survey sampled 209 public and private health facilities [8]. In that survey, the average availability of essential medicines to treat NCDs, across all facilities, was only 15% and regression analyses identified significant associations between availability of these medicines and region, managing authority, facility type, and HIV services [9]. Also, the tools necessary for the diagnosis and/or management of diabetes and cardiovascular disease were available in only 34 and 44% of facilities, respectively. One geographically limited study of overall readiness to manage chronic disease patients in the outpatient units of urban and rural health facilities revealed generally poor readiness to do so [10].

In 2006, in an effort to address the growing NCD burden in the country, the Ugandan Ministry of Health (MOH) created the Programme for the Prevention and Control of NCDs. In a collaboration between that Programme and the Uganda Initiative for Integrated Management of Non-Communicable Diseases, we conducted, in 2013, the first nationwide survey of public sector Ugandan health facilities focused specifically on readiness to provide services for NCDs. By doing so, we were able to conduct a more detailed assessment of NCD services than SARA surveys, which are intended to assess a wide-ranging spectrum of health service delivery readiness. We based our assessment of readiness on the standards set forth by WHO Package of Essential Noncommunicable Disease Interventions for Primary Health Care in Low-Resource Settings (WHO-PEN) [11].

Results of this assessment of the capacity of Ugandan health facilities to address NCDs demonstrate areas of strength as well as significant gaps in the availability of equipment, medicines, and laboratory tests. Although there is variability among the different types of health facilities, none of the facilities surveyed meet the WHO-PEN standards for essential tools and medicines to implement effective NCD interventions [10].

As expected, the RRHs largely fared better than GHs and HCIVs, but a concerning number of facilities lacked equipment, drugs, and standard guidelines for basic and effective NCD prevention and control, based on the WHO-PEN. Particularly concerning at these higher level facilities were the shortcomings in the supply of basic essential technologies. Providers working at health facilities lacking blood pressure machines, measuring tapes, and scales cannot effectively risk-stratify their patients for cardiovascular disease. Without glucose testing strips, providers cannot be expected to deliver adequate care for patients with diabetes.

In addition to basic technologies, all facilities experienced concerning numbers of essential NCD medicines stockouts within the previous year. The recent USAID-funded Securing Ugandans’ Right to Essential Medicines (SURE) [16] is an example of promising work focused on improving access to medicines in the country. However, despite efforts such as SURE, recent evidence demonstrates the continued shortages of, and disparities in, essential medicine availability in Uganda [9]. When medicines and basic technologies are unavailable in the public sector, patients must procure these for themselves in the private sector. However, despite generally better availability, access for patients is still highly restricted by cost. A recent study in Uganda found that 27% of medicines and 32% of diagnostic tests for diabetes and cardiovascular disease were affordable, based on a cost of less than or equal to three days’ wages of a low-level government worker [17]. The importance of improving the supply chain for NCD medicines deserves special mention particularly given that medicine resupply from NMS occurs only every two months. A reduction in stock-outs of NCD medicines will not be achievable without improvements in their supply chain networks. The distribution network for NCD medicines is much less refined than that of HIV medicines which are acquired and distributed though donor-funded mechanisms with sophisticated supply chains [18]. Lessons learned from the successes of HIV supply chains and can be applied to NCD supply chains with minimal investment [19] .

There are a number of concerning features regarding the presence of clinical equipment at the surveyed facilities including high rates of non-functional equipment. Ultrasound machines are often used in the diagnosis and management of chronic kidney disease that is a common outcome of diabetes and hypertension. X-rays are important tests in the management of lung diseases, both communicable (such as tuberculosis) and non-communicable (such as COPD).

The 53 higher level public sector Ugandan health facilities surveyed generally lacked basic NCD prevention and screening services. Though community-based prevention and screening activities are an important component for the promotion of healthy lifestyles and early detection of NCDs [20], health facilities play a vital role in the continuum of services for patients. Despite this, less than half of all health facilities have the essential equipment and tests to screen for NCDs. Many of the facilities do not conduct basic and necessary NCD services, especially at the GH and HCIV level. This is especially concerning given the primary health care focus of these facilities and the likelihood that these facilities see the same patients more regularly than RRHs and therefore have more opportunity to prevent, identify, and manage NCDs. One area of strength in our findings was the availability of individual and group NCD education and nutritional counseling, most of which were offered at a majority of sampled facilities. In addition, 77% of RRHs sampled offered peer NCD support groups. Effective NCD prevention efforts must be more robust and multi-faceted. These should include proven population-level strategies, such as the WHO Best Buys [21], public health awareness campaigs using mass/social media, and community empowerment efforts such as at schools and places of work [22]. But public health campaigns must also be accompanied by individual-level health promotion efforts. At the individual patient level, health promotion education can take many forms [23]. This includes Information, Education, and Communication (IEC) materials, which were offered at only 14.3% of HCIVs and 7.7% of GHs in our study but could also include innovative self-management tools such as illustrated booklets [24] and mobile health platforms that might be leveraged to include NCD-related educational content. Alcohol and tobacco screening was reported to occur in only a small fraction of sampled facilities but these are critical elements of NCD prevention that can easily be performed by healthcare providers and are proven effective [25, 26].

The health facilities in our study also had poor quality assurance methods in place for NCD services. Only 11.5 and 15.4% of GHs in our study utilized diabetes and hypertension guidelines, respectively. A similar study in Tanzania found that 13% of facilities used guidelines for these conditions [27]. Algorithmic approaches to chronic disease management are especially critical in lower-level facilities in which care is provided by non-medical-doctor clinicians. Integrated approaches to comprehensive NCD care, such as WHO HEARTS, which was developed based on WHO-PEN, must incorporate management algorithms that can be easily followed and are locally appropriate and rational for the setting.

This study also highlights the urgent need for tools to allow countries like Uganda to frequently measure and monitor the availability of health service delivery indicators. SARA, an extensive and robust WHO-endorsed survey tool, is meant to be conducted annually. However, the most recent SARA in Uganda was completed in 2014. Such infrequent assessments of health system readiness and service availability create major challenges to the proper planning and effective distribution of limited resources. Health systems face many competing priorities. Not all NCDs are of equal burden and, therefore, not all essential services can, nor should necessarily, be distributed equally. An example in Uganda would be hypertension, which has a prevalence of 26.4% [28] compared to diabetes which has a prevalence of 1.4% [3].This is the basis for WHO-PEN, which itself represents a core set of low-cost, high-impact NCD interventions that can be delivered through a primary care approach [29].

Given that there were three SARA surveys conducted in Uganda between 2012 and 2014, we feel it is important to point out a number of specific ways in which our study adds to the evidence available from those surveys. The 2014 SARA survey was the most similar to our survey in terms of health facilities sampled insofar as it was restricted to HCIV and above. This SARA survey, like our study, inquired about functionality of equipment during the deployment of the survey However, the NCD-related content of this survey was limited to the availability of a few basic pieces of equipment (ie blood pressure machine and stethoscope) and 20 essential medicines. The 2012 and 2013 SARA surveys sampled a wider array of facility types but also were more comprehensive in terms of NCD-related scope. Our assessment of service availability included the presence of many types of personnel, including nutritionists, social workers, and diabetes nurses, among others, and assessed confidence. Comprehensive NCD care requires a team-based approach and continuing education to build confidence and this is not assessed in SARA. SARA assesses medicine availability by spot-checking on pharmacy shelves, while we assessed reported trends in medicine availability over time by inquiring about stock-outs. While neither methodology adequately substitutes for real-time monitoring of availability, we feel these two techniques complement one another and could both be employed in future work. The 2013 SARA reported 80% availability of CVD guidelines, while we found markedly lower availability of these, as discussed earlier. This is illuminating and we feel is related to methodology. While SARA spot-checks for availability of guidelines at facilities, the fact that we asked clinical leaders about the presence of guidelines was likely more indicative of the likelihood that they are actually being used in clinical practice. NCD care requires tools, such as ECG, that are considered by WHO-PEN to be advanced technologies. SARA does not survey for these items. Not only was ECG available in a minority of sampled facilities, but in only 5.6% of all sampled facilities was ECG functional. Future nationally representative assessments of health facility capacity to address NCDs might be bolstered by incorporating some of the methodologies employed by our study.

The local dissemination of our findings led to a guided procurement and distribution process, as described above. Though the number of each of these items was modest, this represents an important appropriation of essential tools by MOH based on the findings of this survey. Documentation of system gaps should always be met by attempts to close those gaps. We believe that the alignment of this needs assessment with MOH priorities and the strategic dissemination process likely aided this process and should serve as a model to other researchers aiming to document health system deficiencies.

This study has a number of limitations. First, the study sample was not randomly selected and did not include all of the GHs or HCIVs. Consequently, the results from this needs assessment may not be fully representative. Second, the “gold standard” of conducting a needs assessment is time-intensive direct observation [12]. However, due to time and funding constraints, each survey was completed in less than one day and the quality and accuracy of the data collected depended on the staff available on the particular day of the study team visit. The levels of knowledge and experience of personnel who helped to complete the survey varied at each facility which may have affected the accuracy of reports. Therefore, future improvements for similar needs assessment surveys would include randomized sampling, standardized data collection methods [30], and time-intensive direct observation with manual spot checks. A more sustainable alternative, though one with substantial associated up-front costs, would be digitization of supply chain management which would allow for improved transparency and data access by individuals within facilities and in health sector planning positions. Finally, in some cases, we observed unexpectedly lower proportions of available essential medicines and technologies in RRHs than in GHs. In our experience, there were likely two explanations for this finding. First, some GHs receive donations from non-governmental organizations rendering them better equipped than their counterparts. Second, facilities differ in their organizational structures and administrative management. Those that are more organized and better managed are likely to be better equipped. Recent analyses of Ugandan SARA data also revealed that some GHs were better supplied than some RRHs [Error! Bookmark not defined.]. Despite these limitations, this needs assessment provided data and information about Ugandan health facilities that was previously unknown and can be considered complementary to other sources of facility readiness data.

This first nationwide research study of the capacity of higher level public sector health facilities in Uganda to provide NCD services highlights many critical gaps. These gaps were addressed, to a limited extent, throught the resulting procurement and distribution process. However, deficiencies such as those of healthcare personnel quantity and training, and access to medicines will continue to limit the potential for improving the quality of NCD service delivery in the country. This study demonstrates the need for Uganda to scale-up low cost, high impact NCD interventions and strengthen the capacity of health personnel to reduce NCD-related disability and death. Improvement of the capacity of health facilities and personnel to effectively detect and manage NCDs can yield significant economic gain for Uganda through a reduction in premature mortality, improved quality of life, and increased productivity [11]. Finally, as the world prepares for the third United Nations General Assembly high-level meeting on NCDs later this year, it is imperative to state that any system changes that seek to fill gaps such as those identified in this study must come in the context of attention to equitable distribution of resources and attention to the socioecomonic determinants of health [21].