Non-communicable diseases (NCDs) are increasingly imposing a greater disease and economic burden globally.¹ For instance, evidence from the 2019 Global Burden of Disease study indicated that nine out of the top 10 drivers of increasing disability-adjusted life years (DALYs) were NCDs.² With the increasing costs associated with accessing healthcare³ especially in developing countries,⁴ NCDs would further burden developing countries and derail the move towards universal health coverage (UHC).

Universal health coverage is a primary target of the Sustainable Development Goal 3, which compels countries to streamline their health systems to ensure that everyone has access to good quality healthcare services that they need without experiencing any financial hardship as a result.^5,6,7 Central to UHC is ensuring that all citizens are protected from financial ruin resulting from accessing healthcare.⁸ However, financial risk protection mechanisms remain hardly adopted in most developing countries. For example, only four countries in sub-Saharan Africa have health insurance coverage greater than 20%.⁹ Consequently, financing of healthcare has largely depended on out-of-pocket payments (OOP) and donors. For instance, 39.6% and 23.4% of Kenya’s total health expenditure in 2016 was composed OOP and donor funding, respectively.¹⁰

The higher dependence on OOP as a financing mechanism results in a higher incidence of catastrophic health expenditure (CHE) (defined as OOP health payments exceeding a certain threshold of household’s capacity to pay or non-subsistence spending)^11,12 and impoverishments,¹³ especially among patients with NCD. Evidence from a study in Kenya indicated that nearly three in five patients with hypertension experienced CHE as a result of direct healthcare costs.¹⁴ In another study among patients with diabetes in Kenya, the incidence of CHE was 75.4%.¹⁵

Despite Kenya’s commitment to achieve UHC by 2022¹⁶ and the prioritisation to deliver NCD services through primary health care (PHC) facilities (level 1–3) – dispensaries and health centres, treatment for NCDs has traditionally been available only in level 4–6 facilities – secondary (county) and tertiary (national) referral hospitals. As a result, very little effort has been put towards community prevention activities, early detection and treatment or improved continuity of care for NCDs; yet these elements are fundamental for successfully addressing the threat of NCDs. It remains unclear how much it costs patients with comorbidities to receive services from these facilities. Although some patient cost of illness studies exists in Kenya, they have focused on single diseases such as diabetes¹⁵ and hypertension¹⁴ separately, yet comorbidities such as hypertension and diabetes often exist. Furthermore, these studies have generally looked at aggregate costs of the diseases at secondary and tertiary facilities.

Against this backdrop, this study aims to examine the cost of outpatient services that patients with two NCDs, diabetes and hypertension, incur to access services at PHC facilities in two Western Kenya counties. Diabetes and hypertension are important modifiable risk factors for cardiovascular disease (CVD), which accounts for over 29% of global deaths.¹⁷ They also have shared pathways of complications and clinical approaches.^18,19 This presents an opportunity for a unified model of care at the PHC level to alleviate a considerable cause of mortality and disability.²⁰ The focus is on outpatient visits as they form the largest share of interaction between patients with NCDs and the health care system and because there is an emphasis on providing these services at PHC facilities that are often closer to patients than secondary facilities. This study aims to generate evidence essential for adequately understanding the cost of delivering these services at PHC facilities. It is our hope that these data will facilitate policy making efforts aimed at increasing access to care of these two conditions at the PHC level in line with the WHO Package of Essential Non-communicable Disease Interventions (WHO PEN) for Primary Care in low-resource settings.²¹

The rising burden of NCDs in developing countries and its resultant economic burden has increasingly attracted the attention of key global health actors and policymakers. Specifically, ensuring NCD healthcare services are affordable is critical in addressing the NCDs burden. This study set out to examine the affordability of outpatient costs two NCD conditions across different levels of health facilities. In general, we found that patients with diabetes and hypertension comorbidity incur higher OOP costs particularly in higher level health facilities and private clinics, but all categories of patients experiencing significant CHE. A study by Subramanian et al.³⁷ established that screening, diagnosis and treatment costs for NCDs are significantly higher in private facilities relative to public facilities, and that these costs were unaffordable to Kenyans. This may be explained by the subsidies in the public sector through government payments. Despite the cost differences, quality studies do not support the claim that the private sector has better outcomes than the public sector, although it is usually more client orientated and performing better in drug availability.^38,39

The direct medical costs across the conditions examined in this study were significantly higher compared to the direct non-medical costs. For example, the total mean annual direct medical cost was KES 7199.00 while the total mean annual direct non-medical cost was KES 1116.00. Of note, prescribed medicines costs were the key cost driver under direct medical costs. While this finding reflects what other cost of illness studies for NCDs have established from other African countries,^40,41,42 this status of affairs is nonetheless worrisome for three reasons. Firstly, it has been established that unaffordability of medicines is the key limitation in the continuity of care for NCDs as patients tend to forego the needed treatment until they have access to funds, resulting in serious medical complications. Secondly, given that more than 20% of the patients in this study obtained their prescribed medication from a government health facility (where the prices of the medicines are subsidised), it is evident that more needs to be done by policy makers to increase budget allocation for healthcare and county managers to ensure lower cost of medication hence increase access and affordability of NCD medicines as medication costs still significantly contribute to financial hardship. For example, the total mean annual medication cost per patient in the overall sample was KES 6528.00 compared to the total mean annual outpatient cost per patient, which was KES 772.00. Further, given that a majority of the patients (66%) sourced their medicines from the RFP, it highlights the inaccessibility of medicines in government facilities because of costs or stock-outs and reinforces the utility of RFP in improving access to NCD medicines especially in rural areas.³⁶ Finally, the high medication costs observed in this study is a concern given that previous studies have established the existence of a wealth gradient in the access to medicines for NCDs in developing countries. A study conducted in eight counties in Kenya, for example, found that poorest households, relative to wealthier households are more likely to have limited access to medication and had to pay higher costs overall to obtain medication for hypertension, diabetes and asthma.⁴³ Moreover, in as much as the finding of this study suggested that most patients obtained their prescribed medication from a government health facility and the RFP, another study conducted in Kenya showed that NCDs medicines were more readily available in the private sector compared to public health facilities.⁴⁴

Transport costs were also another cost driver in this study, highlighting the need to have NCD services closer to the patients. Indeed, transport costs were higher than outpatient costs for NCDs like diabetes and hypertension. Whereas the mean annual outpatient cost was KES 772.00, the mean annual transport cost per patient was KES 1116.00. Previous studies have established that transport cost increases the proportion of NCD patients incurring CHE in Kenya, and transport-related expenses are therefore a likely barrier in seeking healthcare for NCDs.^14,15,45 Another study that evaluated whether health financing reforms that target the poor benefit them identified transport cost as one of the key barriers hindering the poor benefiting from these reforms.⁴⁶ It is worthy to note that although the indirect costs either because of seeking healthcare or hiring caregivers because of the NCDs are lower than the direct costs in this study, they also contributed to the economic burden experienced by households. Elsewhere, a South African study has shown that productivity losses because of long waiting times at health facilities are likely to reduce the demand for NCD healthcare services.⁴⁷

While hypertension and diabetes outpatient services are presently majorly offered in public secondary care facilities and private clinics, our findings suggest that these services as currently distributed are not affordable to most patients, especially those with comorbidity. Overall, 41.83% of the uninsured patients used more than 10% of their income in meeting healthcare costs. Elsewhere, it has been shown that Kenyan households with a member who has a NCD are twice as likely to incur CHE compared to households without a member with a NCD or chronic ailment.⁴⁵ A study comparing CHE by disease area in six countries showed that the share of CHE from NCDs rises with the share of DALYs, defined as years of life lost to disability and death. Non-communicable disease CHE was also more likely to arise from many visits rather than a one-time event, when compared to communicable diseases.⁴⁸ Given the high levels of poverty in Kenya and the low prepayment mechanisms in place to cushion patients from financial hardship because of healthcare payments (80% of Kenyans are not subscribed to any health insurance scheme),⁴⁹ county and national governments need to put strategies to improve the population’s financial risk protection in accessing NCD services. For example, given that most people reside in rural areas in Kenya, dispensaries and health centres should be the focus in increasing service delivery for NCDs as we have shown here that both the direct and indirect costs of care at these primary level facilities are lower than the secondary level of care.

It is noteworthy that this study was undertaken in facilities that had substantial investment in the care cascade for hypertension and diabetes through the PIC4C project. As a result, the cost of care may be lower than that in many other facilities as drugs were made available through the community revolving pharmacies at a lower cost than in private yet in most public settings stock outs are frequent with most patients accessing NCD medication in private. As a result, it is possible that the cost of care is even higher in the rest of the counties.

This study had several limitations. Firstly, as the respondents were recruited in facility chronic care clinics, patients not on care and those that could not afford these care services may be underrepresented. Secondly, because of the recall periods, results may be affected by recall bias. Thirdly, the reliance on an official minimum wage to calculate productivity losses for those without a stated income could have overestimated indirect costs. Lastly, as some respondents did not have income estimates, the CHE estimates should be interpreted with caution. Nonetheless, the data presented are potentially useful inputs in economic evaluation models that require patient costs.

Patients with NCDs still experience substantial costs when accessing healthcare services in Kenya. Evidence from the two Western Kenya counties indicates that NCD comorbid patients incur substantial OOP costs in seeking healthcare, with the poor households spending a higher proportion of their income in receiving health services for NCDs. Measures to ensure affordability of outpatient NCD services should be urgently implemented while concerting prevention and screening efforts to reduce NCDs burden in the country. Provision of NCD services at PHC facilities could ensure achievement of both. Estimates from our study can be used to inform the inclusion of NCD services in the rollout of the UHC scheme in Kenya.