Menu
Menu
Menu
Menu
Background: Direct and time costs of accessing and using health care may limit health care access, affect welfare loss, and lead to catastrophic spending especially among poorest households. To date, limited attention has been given to time and transport costs and how these costs are distributed across patients, facility and service types especially in poor settings. We aimed to fill this knowledge gap. Methods: We used data from 1407 patients in 150 facilities in Tanzania. Data were collected in January 2012 through patient exit-interviews. All costs were disaggregated across patients, facility and service types. Data were analysed descriptively by using means, medians and equity measures like equity gap, ratio and concentration index. Results: 71% of patients, especially the poorest and rural patients, accessed care on foot. The average travel time and cost were 30 minutes and 0.41USD respectively. The average waiting time and consultation time were 47 min and 13 min respectively. The average medical cost was 0.23 USD but only18% of patients paid for health care. The poorest and rural patients faced substantial time burden to access health care (travel and waiting) but incurred less transport and medical costs compared to their counterparts. The consultation time was similar across patients. Patients spent more time travelling to public facilities and dispensaries while incurring less transport cost than accessing other facility types, but waiting and consultation time was similar across facility types. Patients paid less amount in public than in private facilities. Postnatal care and vaccination clients spent less waiting and consultation time and paid less medical cost than antenatal care clients. Conclusions: Our findings reinforce the need for a greater investment in primary health care to reduce access barriers and cost burdens especially among the worse-offs. Facility’s construction and renovation and increased supply of healthcare workers and medical commodities are potential initiatives to consider. Other initiatives may need a multi-sectoral collaboration.
Tanzania is a lower middle-income country in East Africa with an estimated population of around 56 million people in 2016 [36]. Tanzania has 31 regions and most (70%) inhabitants are residing rural areas. Tanzania has made progress on child survival, with little improvement in maternal health [37, 38] and this is regarding the Millennium Development Goals (MDGs) of reducing by two-thirds of child mortality and by three-quarters maternal mortality ratio. In particular, over the past 15 years from 1999 to 2015/16 in Tanzania, the infant and under-5 mortality rates have declined from 99 deaths to 43 deaths per 1000 live births and from 147 to 67 deaths per 1000 live births, respectively [37]. The maternal mortality ratio also declined from 578 deaths to 454 deaths per 100,000 live births in 2004/5, before raising up to 556 deaths in 2015/16 [37, 39]. Access to one antenatal care (ANC) is almost universal, but there remains relatively low coverage of at least four ANC visits (51%), institutional delivery (63%) and postnatal care (PNC) (33%) [37]. The use of maternity services shows a marked imbalance along the continuum of care as reported elsewhere [40–42]. Also, 75% of Tanzanian children age 12–23 months received all basic vaccinations [37]. The health system in Tanzania involves a predominance of public sector facilities, followed by faith-based providers, and a limited number of private-for-profit providers. The public health system has a hierarchical administrative structure, with a referral structure such that dispensaries, health centres, and district hospitals provide primary health care (PHC) services. A dispensary is supposed to serve at least one village, and a ward for a health centre [43]. Tanzania implemented a Primary Health Service Development Programme (2007–2017) to improve access to basic health care service by rehabilitating and constructing at least one dispensary per village and a health centre per each ward countrywide [43]. The health financing system in Tanzania is highly fragmented with many sources including general taxation (34%), donor support (36%), out-of-pocket payments (22%), and health insurance contributions (8%) [44]. In 2018, the health sector review revealed that 33% of Tanzanians are covered by health insurance, which include 8% by National Health Insurance Fund (NHIF) for public servants mainly, 25% by improved Community Health Fund (iCHF) for people working in informal sector, and 1% by private insurance and Social Health Insurance Benefit (SHIB) [45]. The coverage of health insurance is still low, which exposes many Tanzanians to financial risks due to direct health care payments. Despite exemption and waiver policies in Tanzania which aim to protect poor and vulnerable groups (e.g., pregnant women, children, and elders) [46, 47], the enforcement of these policies is weak [29, 48]. Also, the existing health insurance schemes in Tanzania only cover medical expenses at facilities, but do not compensate patients for travel and time costs incurred when accessing care. Data were collected from a cross-sectional survey of patients from three regions (Pwani, Morogoro and Lindi) in Tanzania. All seven districts of Pwani region and four districts from Morogoro and Lindi region were included. This study was part of the large baseline survey of an impact evaluation of a pay for performance (P4P) programme in Pwani region [49, 50]. The evaluation study used Morogoro and Lindi as comparison regions. A sample of 75 facilities from Pwani region were considered and the same number from comparison districts, including hospitals (n = 6), health centres (n = 16) and dispensaries (n = 53) in each arm. Comparison facilities had similar levels of outpatient care visits and staffing levels to intervention facilities. In total,150 public and private health facilities (12 hospitals, 32 health centres and 106 dispensaries) were surveyed (82% were public facilities). Data were collected through patient exit-interviews to a maximum sample of 10 clients/ patients per facility between January and February 2012. Clients were approached upon arrival at the facility, asked a series of screening questions to check their eligibility. Eligible respondents included those resided in that area for at least 6 months, aged at least 18 years, and seeking care for one of the following four services: (i) ANC, (ii) child vaccination for under 1 year, (iii) PNC follow-up for mothers/ babies 2 months after birth, and (iv) check-up for fever, cough and diarrhoea for women/ under 5 children. Thus, respondents included pregnant women, mothers or care givers who brought under 5 children to the facility. Prior to the interview, all eligible clients were asked for their consent to participate in the survey after exiting the consultation room. The exit-interview tool was adapted from the World Bank Impact Evaluation Toolkit [51], which measured a range of quality-of-care indicators including patient satisfaction/ experience of care, and costs of accessing and utilising health services. The exit-interviews also captured information on household background characteristics (e.g., ownership of assets and housing characteristics) that were used to assess the household’s socioeconomic status. All the interviews were conducted in Swahili language. A tool was pre-tested for consistency, relevance, and clarity before the actual survey. The outcome of interest includes time costs as well as transport and medical costs. Time costs were estimated in minutes associated with traveling to and waiting or receiving consultation at the facility. Transport and medical costs were measured in local currency, Tanzanian shilling (TZS), and then converted into US dollar (USD) using the approximate exchange rate during the survey in 2012 (1 USD equal 1600 TZS). All costs were estimated based on patient recall. Transport and time costs of travelling were measured for one-way journey to the health facility. We ‘multiplied by two’ for simplicity to account for a return trip, since patients on their return sometime pass via markets or to other social activities as previously reported in Tanzania [34]. In order to avoid overestimation, the one-way journey is preferred. For robustness check, however, we presented the estimates for both one and two-way travel costs. We examined the distribution of time and direct costs by two dimensions of equity – (i) place of residence (rural/urban) and (ii) household’s socioeconomic status (quintiles). The rural-urban dimension was considered to reflect the remoteness and how facilities are scattered, which has important implications for transport costs and travel time; while the socioeconomic status was included to measure the households’ living standard as a proxy of ability to pay. Household socioeconomic status was assessed through a wealth index based on household characteristics and asset ownership derived using principal component analysis based on 42 items (Appendix Table 5 & 6) [52, 53]. Patients were ranked by wealth scores from poorest (low score) to least poor (high score), and classified into five equal-sized quintiles. We first described the mean and median costs by patient socioeconomic status and residence. The equity analyses proceeded by using three measures of inequality –an absolute measure (the gap) and two relative measures (the ratio and the concentration index) [15, 54]. The equity gap was measured as the difference in costs between patient subgroups, while the equity ratio was measured as the ratio of costs between patient subgroups. Specifically, both equity gap and equity ratio were calculated between poorest and least poor patients, as well as between rural and urban patients. When comparing the poorest and least poor patients, for example, a positive (negative) gap and a ratio greater (less) than one defines high-cost burdens among the poorest (least poor), respectively. A gap of zero or a ratio of one defines an equal distribution in costs. We also used t-tests to assess whether the gaps were significantly different from zero. In addition, we computed the concentration index (CI) to quantify the degree of socioeconomic-related inequality in cost burdens of seeking and receiving health care. The CI was computed on a ranking variable of household socioeconomic status as shown in Eq. (1) [15, 55]. where yi is the cost variable of the ith patient; Ri is the fractional rank of the ith patient (in terms of households’ socioeconomic status, with lower fractions for poorest and larger fractions for richest); μ is the average cost and cov denotes the covariance. The CI ranges between [− 1 and + 1], whereby zero indicate equality between socioeconomic status subgroups, while negative and positive values indicate that poorest have high-cost burdens and low-cost burdens, respectively. We also tested whether the CIs were significantly different from zero. As a robustness check, our analysis was also restricted to public facilities (82%) as these facilities are supposed to offer free MCH services in Tanzania. All analyses were performed using STATA version 16.
N/A
