Demand-side Factors Affecting Health Seeking Behavior in Ghana

GUJHS. 2008 Apr; Vol. 5, No. 1.

Stephen Russell
Georgetown University
Dodowa Health Research Centre, Ghana


Introduction
Since 1997, Ghana has focused on increasing access to health care. In its poverty reduction strategy, Ghana plans to increase the percentage of citizens utilizing qualified health personnel when sick from 30% to 50% (1). This goal remains elusive because of poverty in rural communities, lack of skilled health workers, and lack of quality health providers. Over one third of the country lives over five kilometers away from the nearest health provider, further hampering access for the rural population (2). Even for individuals living closer to health providers, financial and cultural barriers sometimes make health providers inaccessible.  Some individuals use informal health care such as traditional healers, drug peddlers, and drug stores to save time and money. There is inadequate knowledge of what attracts individuals to seek care at orthodox health providers, and even less knowledge about what keeps individuals away from those health providers (3). This investigation of demand-side barriers affecting health seeking behavior in Ghana explores some of these barriers to inform Ghana’s health system on appropriate interventions.

In the complex decision to seek medical attention, cost is a significant factor. This study focuses on costs, both direct and indirect, involved in a visit by an individual to a health provider. Costs include cost of care, cost of transportation, time spent at the health provider, and inability to work while at the health provider. G.J.M van den Boom estimates that consulting a doctor takes on average four hours, and costs eight times more than self-medication in rural Ghana. He also estimates that one trip to the doctor may cost a family 1/3 of their monthly income (2). There staggering numbers indicate that cost is an important factor affecting health seeking behavior of rural Ghanaians.

Other main focuses of this study include the roles that distance from health providers and perceived quality of health provider options plays in the household’s decision of health provider. Recent research has brought to light the fact that physical need is not as significant as predisposing, enabling, and restrictive factors in determining health utilization (4). Different studies have addressed this issue, with varying results. The restrictive barriers preventing individuals from seeking health care differ by country and region. This study identifies significant factors limiting health seeking behavior in Prampram area council, a sub-district in southern Ghana. Understanding the demand side of health care will better inform the supply side in the hopes of providing greater access to and utilization of health services in Ghana.

Research Objectives

  1. To examine the effect that distance from health providers has on health seeking behavior of household heads
  2. To quantify direct and indirect costs of obtaining health services
  3. To study the effect that perceived quality of different health providers has on the health seeking behavior of household heads

Methodology
Target Communities
The Ghana Health System is organized into 110 districts with a hierarchical structure from the Ministry of Health to the community level. The communities involved in this study are located in Prampram sub-district, which falls under the auspices of Dangme West district and Greater Accra region. The communities involved in the household questionnaire are Tetteh Wayo, Abia, Akokokrom, and Yodue Kope. They are rural communities, lying along the southern coast of Ghana.

These communities were targeted for their small size, as well as their relative distance from both private and public health facilities. Individuals in these communities have access to a public clinic, a private clinic, prayer camps, traditional healers, drug stores, and drug peddlers as potential health providers. Each community is located between two and four kilometers from Prampram Health Centre and Ebenezer Clinic, the two largest health facilities in the area. Three of the communities- Tetteh Wayo, Akokokrom, and Abia- are located between two and four kilometers from New Ningo CHPS, a large health facility the opposite way down the main road. Communities were chosen using computerized Geographical Information Systems (GIS).  See Figure 1 in Appendix for the computer generated model, with circles of radius 2 and 4 kilometers from Ebenezer Clinic, Prampram Health Centre, and New Ningo CHPS.

Sampling
Once the communities had been selected for their proximity to health facilities, Dodowa Demographic Surveillance System (DDSS) was used to target houses, households, and household heads in those communities. Houses are physical structures holding one of more households. The household is a family; more than one household can live in the same house. The number of household surveys completed was 103.

Data Collection Tools
The household questionnaire, the bulk of the data collection, is a compilation of 36 open and closed ended questions. Open-ended questions were used to collect qualitative data, which was later coded for use in data analysis. 103 surveys were completed over a four day period, from 6 November 2007 to 9 November 2007.

In addition to household surveys, exit interviews were conducted at Prampram Health Centre and Ebenezer Clinic. The Prampram Health Centre is the government health facility while Ebenezer Clinic is a private institution. Exit Interviews contain 17 open and closed ended questions. 57 exit interview were completed between 7 November 2007 and 9 November 2007.

Training of Field Staff and Pre-Test
The recruitment and written test of field workers was completed on 31 October 2007. A group of seven was trimmed to three who participated in the study. Training commenced on 1 November 2007 and completed on 3 November 2007. Training included a general orientation to the Dodowa Health Research Centre and an introduction to fieldwork. The household questionnaires and exit interviews were explained, broken down into the native language, and practiced. At the end of training, all three field workers completed a competence quiz to ensure quality in the field. Training was conducted by Mr. Emmanuel Tetteh Amano, Ms. Lucy Yevoo, and myself. The former two assisted me as field supervisors.

A pre-test was conducted in Dodowa town near the Dodowa clinic. Each field worker completed three exit interviews from Dodowa Clinic and practiced the household questionnaire on three others. Based on the results of these surveys the data collection tools were reassessed and corrected.

Community Entry
I was introduced to a contact person for all four communities on 6 November 2007. He gave his permission for the study and invited me to interview in each communities. On 6 November 2007 I met with the heads of both the Prampram Health Centre and Ebenezer Clinic, obtaining consent to administer exit interviews on their premises.

Data Analysis
Data was entered into EpiData. It was then analyzed using Excel and SPSS for frequencies, charts, tables, cross tabulations, and single-mean T tests. A 95% confidence interval was used to determine significant relationships between data.

Results
Demographics
The communities are within six kilometers of each other and between two and four kilometers of the nearest health facility. The questionnaires were given to heads of households or household representatives. In most cases where the household head was not available, the respondent is the spouse of the household head. Respondent’s ages range from 16 to 80, with seven individuals unsure of their age. The greatest frequencies are between 20 and 40, shown in the graph below.

Number of respondents varies by community. The communities are relatively small; Abia has a population of 689, Akokokrom has 498 residents, Yodue Kope has 395 residents, and Tetteh Wayo has 88 residents. The number of respondents by community is represented in the graph below.

Females account for 69 of the 103 respondents, or 67%; 34 Males were interviewed. The majority of those interviewed were of the Ga-Dangme ethnic group (64%), the second most popular group being Ewe (32%). Of the Ewe respondents, thirty reside in Yodue Kope and 3 live in all other communities combined. Abia is predominantly Ga-Dangme, with 41 out of the 42 respondents. Akokokrom and Tetteh Wayo also have high percentages of Ga-Dangme ethnic group, 78 and 80 percent respectively.

Many of the respondents are married (63%) or are in co-habitation with their spouses (19%).  The distribution of marital status is shown below.
Educational Status showed a bit more variation than some of the other demographic statistics, with high frequencies for no education (38%), primary school education (29%), and Junior Secondary School education (26%). Other education levels include Senior Secondary School (5%) and Tertiary education (2%).

Respondent’s occupations were diversely distributed, as well. The predominant main occupations were trader (39%), farmer (20%), and fisherman (12%). The full range of responses for main occupation with frequencies is shown below.

Health Seeking Behavior and Health Provider Choices
The decision to seek out health services is a complex decision. For any study into the reasons underlying health provider choices and health seeking behavior, it is important to identify the health providers that individuals frequent. Individuals in Tetteh Wayo, Abia, Akokokrom, and Yodue Kope have access to multiple health providers including Prampram Health Centre, Ebenezer Clinic, drugs stores, drug peddlers, prayer camps, traditional healers, and others. Prampram Health Centre and Ebenezer Clinic are the largest and most popular health providers; Prampram is a government clinic and Ebenezer is a private clinic. The graph below show the frequency of respondent’s who usually go to the various health providers.

For most respondents, the health provider that they usually go to is the closest to their house. Only 27% of respondent’s say that a different health provider is closer to their house. Also, 80% of the respondents go to the cheapest nearby health provider, indicating that individuals regularly go to the nearest and cheapest options for health care. Respondents cited New Ningo CHPS, drug stores, drug peddlers, and prayer camps as providers either cheaper or closer to their houses that they chose not to go to. Since the percentage of individuals seeking the closest and cheapest health providers is not 100%, it seems that more than proximity and cost of health providers affects individual’s health provider preferences.

Indeed, respondents tend to shop around when it comes to health providers. 68% of respondents have been to a health provider other than the provider that they usually go to. Providers on this list include Ebenezer Clinic (35%), Prampram Health Centre (29%), drug store (20%), and the drug peddler (9%). No respondents have ever been to a traditional healer for health care.

An analysis of respondent’s health provider preferences reveals no difference between communities. This belies the chance that there is some community-based social impetus for one health provider over another. Nearly all respondents, furthermore, believe that other members of their community go to the same health provider as they do (89%).

The choice to seek health care is more than just which health provider to seek; there is a timing aspect as well. 28% of respondents go to their health provider the first day that signs of illness appear. Another 28% wait for the next day, while 5% wait for two days and 39% wait for three or more days. This decision also has multiple factors affecting it, such as severity of disease or operating hours of the health provider. There is a positive relationship between quality of the road and promptness in seeking health care. This has important implications for Ghana’s health system in general and health infrastructure in particular. The most common form of transportation to the health providers is walking (53%), followed by taxi (34%), and tro-tro (10%). A tro-tro is a large passenger van that works like a taxi but fits more passengers at a cheaper price.

There is an inverse relationship between the promptness of care seeking behavior by health provider. For instance, individuals who seek care primarily at Prampram Health Centre seek care earlier after the first signs of illness than respondents who choose Ebenezer Clinic. The same relationship applies to respondents that go to the drug store. One suspected reason is that the high cost of Ebenezer Clinic discourages respondents from seeking health services until the illness becomes more serious. The quicker that an illness is treated, the better the outcome; this statistic, therefore, has important implications for health providers and the health system writ large.

Cost
(Note: 10,000 cedis ~ 1.00 USD)
The cost of seeking health care includes more than the cost of consultation and medicine; it involves hidden costs. In economics these are often referred to as opportunity costs. My study focuses on cost of treatment, time spent seeking health care, and the loss of income from not working. All or some of these costs may weigh into an individual’s health seeking behavior.

The most obvious cost, and the easiest to measure, is direct cost of care. Respondent’s costs, the last time that they visited their health provider, ranged from zero to 1,000,000 cedis (~100.00 USD). Respondents spent an average of 109,425.70 cedis (~10.94 USD) the last time that they sought care at a health provider. Between health providers, patients paid an average of 134,914.3 cedis at Ebenezer Clinic and 116,804.4 cedis at Prampram Health Centre. Respondents spent significantly more at Ebenezer Clinic than at Prampram Health Centre or the drug store.

Boxplot of Health Provider vs. Cost of Care

Based on exit interviews, patients spend less at both Ebenezer Clinic and Prampram Health Centre than household questionnaire respondents. Patients at Ebenezer Clinic spend 70,000 on average, but only 36,000 at Prampram Health Centre. These numbers, though based on fewer surveys, may show that respondents from rural communities wait for more serious illnesses before going to health providers, resulting in the higher costs.

Another important direct cost is the cost of transport to the health provider. Most respondents walk to their health provider (53%), while others take taxi (33%) or tro-tro (10%). The average cost of a tro-tro ride is 6,060 cedis and 17,485 cedis for taxi. The average cost of those who took transportation other than walking was 15,000 cedis. This number is larger than the average for exit interview patients, who paid only 8,000 cedis on average for transport. Of the exit interview patients, a higher percentage of respondents (70%) walk to the health provider. This indicates that individuals frequenting the health clinics live close to the health providers, and individuals from rural communities spend more money to get to health providers and spend significantly more at the health providers.

Wait time at the health provider is a key opportunity cost. 51% of respondents are treated within the first hour, and 33% in the second hour. This number depends, however, on the health provider that the respondent goes to. Ebenezer Clinic and drug stores have significantly shorter wait times than Prampram Health Centre. The graph below shows the wait time of respondents by health provider. Calculated from exit interviews, the average wait time at Ebenezer Clinic is 39 minutes, while the average wait time at Prampram Health Centre is 2 hours 13 minutes. These numbers are consistent with the wait time perceived by household respondents.

Wait Time by Health Provider

Another significant opportunity cost incurred by the decision to go to a health provider is the loss of income from a day’s work missed. With the combined travel time to and from the clinic, as well as waiting time, a trip to a health provider may negate a whole day’s worth of work. Many individuals will work through illness, so the only time that they cease working is when they go to the health provider. Of the respondents, 56% have no one to take care of their work while they go to the health provider. Of the 46% that do have someone take care of their work, seven pay for an individual to work for them, and others give gifts or food in appreciation.

29% of respondents do not have anyone to take care of the household chores while they are at the health provider. Of the 71% that do, it is often a family member who is sometimes paid or given gifts for the service. These hidden costs of going to a health provider can be powerful disincentives for seeking health care.

Distance of Health Provider
An important factor in the decision to seek health services is the time that a trip to the health provider takes.  Each community is less than four kilometers away from every health provider. The time it takes to get to the health provider varies by community; it takes individuals from Yodue Kope longer to get to the health providers because of the poor quality of the road leading to the health providers. For all communities except Yodue Kope the road to Prampram town, where most of the health providers are located, is the same.

Respondents need little time to get to the health providers; 93% of respondents were within one hour of the health provider, it took longer than an hour for only 7% of respondents.  Most of these individuals, however, walk to the health provider, which may damage their already fragile health status.
Health provider choices did not vary by community. Distance from health facilities may come into play for communities farther from viable health care options, but each of the communities studied has access to several health provider options, and distance was not the most important factor in health seeking decisions.

A subtler look at the role of distance from health provider, however, sheds some light on the fact that distance may be a significant factor in the health decision process. 70% of patients in exit interviews at Prampram Health Centre and Ebenezer Clinic walked to the facility. Based on household surveys, only 53% walk to the health provider. Also, exit interviews reveal that patients pay significantly less than household respondents in the rural communities. This indicates that individuals frequenting the health clinics live closer to the health providers than the communities studied.

This apparent distance factor prevents individuals in rural communities from going to the health provider as often as individuals who are closer. Based on the fact that patients in exit interviews pay significantly more than household respondents pay, household respondents may wait longer and wait for more serious illness before going to orthodox health providers. This is a relationship that deserves more attention, as this analysis can not conclusively highlight distance as the determining factor in this apparent negative relationship between distance from health provider and health seeking behavior.

Perceived Quality of Health Provider Options
Cost plays a major role for individuals deciding which health provider to go to, while distance from the health provider is a less important proxy. Perceived quality is another major issue; it follows reason that individuals will only go to a health provider that they perceive is high quality. The fact that 68% of respondents have been to health providers other than the one they usually go to is further evidence that quality is an important factor.

Respondents feel that when they go to their health providers they are cured of illness. Only one respondent is never healed when he/she goes to the health provider and 25% are sometimes cured. 74% claim that they are cured every time they go to the health provider. Between providers, the percentage of those who are cured every time is 70% for Prampram Health Centre, 83% for Ebenezer Clinic, and 44% for the drug store.

Respondents are well treated by the staff at health providers. 82% feel that they are “very well” treated, whereas only 5% say they are treated “okay” and only 1% is treated “badly”. Based on exit interviews at Prampram Health Centre and Ebenezer Clinic, 12% of respondents choose their health provider over other nearby providers because of the good treatment by health workers. Exit interviews also indicate that 32% of patients respond that kind treatment by health workers is the best thing about their visit to the health provider.

Most respondents recommend their health provider to others, another measure of patient satisfaction. 86% of respondents recommend their health provider to others, a high percentage considering the number of other available health providers.

On most measures of patient satisfaction and overall health provider quality in both the household questionnaire and the exit interviews, Ebenezer Clinic exceeded all other health providers. Their perceived rate of success (83%) is highest among frequented health providers, as is their percentage of patients who recommend Ebenezer to others (100%). In exit interviews, all 28 patients at Ebenezer Clinic were treated within the first hour, with an average wait time of 38.6 minutes. Respondents cite adequate provision of medicines, cleanliness, health worker attitudes, morning prayers, and short wait time as the reasons that they choose Ebenezer over other health providers.

One of the drawbacks to Ebenezer Clinic is significantly higher cost than other health providers. Respondents on exit interviews recommend that Ebenezer Clinic expand its facilities and improve health equality. Some patients cite preferential treatment of some patients over those with government issued health insurance. Patients at Ebenezer Clinic, on average, wait significantly longer to go to the clinic after the first signs of illness; a problem which has major health implications for those individuals.

Respondents who go to Prampram Health Centre also indicate a high success rate (70%). Respondents also recommend Prampram Health Centre to others at a very high rate (82%), although lower than the rate of recommendation for Ebenezer Clinic (95%). Respondents who recommend Prampram Health Centre do so for different reasons than at Ebenezer Clinic. Patients choose Prampram Health Centre for its low cost, distance from house, health worker attitudes, and inability of other health providers (e.g. prayer camps) to deal with some illness (e.g. malaria). Respondents note health worker attitudes, prevention programs, and past success as the best things about Prampram Health Centre. Patients who usually go to Prampram Health Centre do so early after the first signs of sickness, usually the same day or the next after the first signs of illness, improving their recovery time.

A drawback to going to Prampram Health Centre is significantly longer wait time. Based on exit interviews, patients wait an average of 2 hours and 11 minutes before seeing a doctor. Patients recommend providing adequate medicine, expanding facilities, improving operating hours, and shortening wait time to improve Prampram Health Centre.

Why Respondents Choose Health Provider Over Others They Have Been To

Discussion
It is important to study demand-side barriers to health services in rural communities where they are more prevalent. The communities are among the poorest in the area, so they have access to orthodox health providers less often due to cost, distance, perceived quality of options, and other restrictive factors. Among the restrictive factors studied, cost and quality of health provider options are the most important factors in health seeking behavior of respondents. Education level is a significant enabling factor, but much work has been done on this relationship and it is not explored in this study. Most of the analysis was done between communities or primary health provider to establish health seeking behaviors by geographical location as well as choice health provider. Demand side barriers including distance, cost, and quality of care affect respondent’s health seeking behavior in Prampram sub-district.
The most important health decision that households make is the health provider. A household’s health provider is defined as the health provider that they usually go to when someone in the family is ill. Each community is located relatively close to Prampram town, where there are multiple health providers. The most popular health provider choices are Prampram Health Centre, the nearby government health centre, and Ebenezer Clinic, a private clinic. As these are the two most popular health providers, additional data collection was collected from exit interviews at each health facility. Households also identified drug store as a primary provider. Almost all respondents have been to a different health provider, however, which indicates that respondents may make choice of health provider relative to cost, quality, and other factors. For this study, there is no measure of health literacy and level of self-medication, which is an alternative that demands further exploration.

Since health-seeking behavior does not vary by community but distance does, it indicates that distance may not be a disincentive for respondents to seek health care. The communities do not vary in how long after onset of illness they go to the health provider, time it takes to get to the health provider, and choice of health provider.

The only distance factor that is significant is for Yodue Kope.  While Yodue Kope is not farther from health providers than Abia, Tetteh Wayo, or Akokokrom, respondents from Yodue Kope wait longer after onset of illness before seeking care at a health provider. It also takes them significantly longer to get to the health provider. This can be explained by road quality, which is significantly worse than for roads to the health providers from the other communities.  Comparing health seeking behavior of household respondents and exit interview respondents reveals a potential distance factor in health seeking behavior. 53% of household respondents walk to the health provider, while 70% of exit interview respondents walked to the health provider. This indicates that those frequenting the health providers are closer to the health providers. Household respondents, therefore, may make less use of health providers and distance may be the factor preventing them from equal access.

Since health providers were so clustered in Prampram town, it is not surprisingly that distance was not a conclusive disincentive. Results for cost were more interesting; cost of care varies significantly between health providers. Informal health providers such as the drug store or the prayer camp are cheaper options than the established health facilities Prampram Health Centre and Ebenezer Clinic. The cost difference between Prampram Health Centre and the drug store, however, is not significant but respondents choose Prampram Health centre for adequate medicine and better treatment. Respondents who go to the private Ebenezer Clinic pay significantly more for care than respondents at the public Prampram Health Centre and any other health provider. The fact that 38% of respondents go to Ebenezer Clinic despite the higher costs indicates either that those individuals are wealthier or factors other than cost are significant in health seeking behavior.

While relative cost differences between health providers are important, cost of care remains a resounding factor for most respondents, even those who choose a private health provider. 80% of respondents say that the health provider that they go to is the cheapest nearby option. A preference for cheap care is obvious and unequivocal.

Time spent visiting a health provider is another cost for respondents. Those who choose nearby health providers such as the drug store and prayer camps note its low cost and short wait time as reasons. Respondents in the household questionnaire and exit interview alike praise Ebenezer Clinic for its short wait time and optimal operating hours and criticize Prampram Health Centre for long wait times and unpredictable operating hours. Based on responses from exit interviews, patients wait about three times longer at Prampram Health Centre than Ebenezer Clinic to see a doctor. At drug stores, the wait time is significantly shorter than at Prampram Health Centre and Ebenezer Clinic. Respondents who can’t afford to waste time may be attracted to informal options such as the drug store. Any extra wait time pushes up the total cost of seeking health care for respondents, as they lose income from work missed, and their condition deteriorates the longer they wait for attention.

Another cost incurred in going to the health provider is the cost of transport to the health provider. Patients choose between walking, tro-tro, and taxi. Among those options, taxi is the most expensive but the fastest and most reliable. Walking is a negative option for any individual who is ill. Physical activity under a hot sun can worsen illness, especially when patients have to walk an hour to the health provider. An average taxi ride costs about 17,000 cedis while an average tro-tro ride cost respondents 6,000 cedis. Based on averages of cost of care at the health providers, a two-way taxi ride amounts to 20% of the cost of care, while a two-way tro-tro ride is 11% of the cost of care. These extra costs push up the overall cost of seeking health care at health providers, further discouraging rural Ghanaians from seeking prompt care.

The fact that household respondents from Tetteh Wayo, Yodue Kope, Abia, and Akokokrom pay more on average to get to health providers and pay more at the health providers than those interviewed in exit interviews is an important finding. Individuals who frequent Prampram Health Centre and Ebenezer Clinic live closer than those living in the communities chosen for household questionnaires. One result of this relationship may be that rural communities self-medicate more often. Household respondents may only go to health providers when their illness is more serious, an explanation of the significantly higher costs. Further studies may reveal an even larger gap between health costs of individuals closer to health providers and those farther away.

Beyond what respondents pay getting to the health facility and at the health facility, income is lost from their inability to work while at the health provider. 56% of individuals have no one other than themselves to run their business. Some pay others to run their business, some give gifts, and some provide food in exchange. Individuals may work through sickness instead of going to health providers because of the actual and time costs, as well as their loss of income from a day’s work lost.

Wait time at the health provider before seeing a doctor is both a cost and a quality issue. Based on exit interviews, individuals waited at average of forty minutes at the Ebenezer Clinic, but patients at the Prampram Health Centre waited over two hours before seeing a doctor. Besides being a nuisance, such a long wait may deter patients from Prampram Health Centre to less effective health providers in the same price range. Since forty-seven respondents usually go to Prampram Health Centre and forty-two usually go to Ebenezer Clinic, it is incorrect to make the argument that Prampram Health Centre takes longer because of an overload of patients. Similar numbers of patients frequent both clinics and they have comparable numbers of health workers, so there must be a quality issue that is causing these significantly different wait times. In defense of Prampram Health Centre, some of the exit interviews were of women who had stayed for a maternity class, so their time at the clinic would be longer than an individual coming in for a checkup. Still, the vast differences in wait time suggest a quality difference.

On other measures of quality, however, Prampram Health Centre meets Ebenezer Clinic. Respondents who frequent the clinics are generally cured of their sickness; other health providers such as the drug store have lower success rates. These are, however, only perceived success rates; a cohort study is required to determine actual success rates.

The study reveals little data about the quality of other health providers such as traditional healers, drug peddlers, and prayer camps. Few respondents frequent these health care options, although some have been to them before. 20% of respondents have gone to the drug store before, 9% have bought drugs from the drug peddler, but no respondent had ever been to a traditional healer. Indeed, some exit interviews respondents say they usually go to other health providers, but they understand that those health providers have limits on the types of illnesses they can treat. They come to orthodox health providers when the illness is too serious for other health providers. Based on exit interviews, 31% of respondents chose Prampram Health Centre over other health care providers because those health providers can’t cure all treatments. Patients noted that traditional healers and prayer camps do not deliver pregnant women and have no cure for malaria.

Quality of health providers is a significantly important factor in the health decision making process. Respondents chose their health provider over cheaper options for reasons including adequate provision of medicine (37%), health worker attitudes (32%), proper facilities (11%), and wait time (11%). These qualities of some health providers may sometimes trump the lower cost of other health providers.

Respondents who visit Prampram Health Centre as their primary health provider mentioned cost (38%), adequate medicine (13%), past success (9%), health worker attitudes (4%), and distance from house (2%) as reasons that they choose Prampram Health Centre over other health providers that they’ve been too. Exit interviews revealed that respondents enjoyed the health worker attitudes, short wait time, and prevention programs at Prampram Health Centre. While these numbers indicate a quality health facility, some of these numbers were significantly less than the perceived quality of Ebenezer Clinic. The wait time in particular is skewed; the wait time at Prampram Health Centre is longer than household respondents perceive. One of the best things about Prampram Health Centre, however, is their unique maternity education program. Even those who go to Ebenezer Clinic say that they go to Prampram Health Centre for the maternal health programs. While Prampram Health Centre scored well on some quality measures, Ebenezer fair even better.

Respondents who go to Ebenezer Clinic mention some of the same reasons as respondents who go to Prampram Health Centre. Those who go to Ebenezer Clinic like the health worker attitudes, adequate medicines, and short wait time. They also enjoyed morning prayers and the cleanliness of the facility. Respondents from exit interviews suggest that Ebenezer Clinic keep up the good work, treat patients equally regardless of health insurance status, and expand their facilities. There was an interesting plea for patients from Ebenezer Clinic to expand to include a maternity ward, highlighting a lack of maternal care.

Regardless of health provider choice, respondents generally go to their health provider quickly when someone in the household is sick. Respondents who go to Ebenezer Clinic, however, wait significantly longer before seeking care. Waiting three or more days before going to the clinic, respondents increase the severity and complexity of disease. This may be an indication that Ebenezer Clinic and its patients would benefit from more health promotion and disease prevention. This could be in the form of pamphlets in the waiting rooms or structured health education classes like those at Prampram Health Centre.

Respondents who go to Prampram Health and the drug stores first, on the other hand, generally do so shortly after onset of illness. By going the same day or the next day after onset of illness, respondents have a better chance of quickly overcoming the illness. Wait time at Prampram Health Centre, however, is significantly longer than Ebenezer Clinic and the drug store. While the drug store has a shorter wait time, their success rates are also significantly lower than either of the two clinics.

While Prampram Health Centre and Ebenezer Clinic differed in cost, there was no difference in the percentage of those with health insurance visiting the clinics. There is also no different in health seeking behavior of individuals with health insurance. There seems to be some misinformation in the system; some respondents in exit interviews say that they go to Prampram Health Centre because they accept the green card, the national health insurance program. Individuals with the green card, however, are accepted at Ebenezer Clinic. A few of these respondents note in exit interviews that Ebenezer Clinic discriminates against holders of the green card; two respondents allege that patients with the green card are seen after those who pay out of pocket. Although only a few mention this issue, it may grow in the next few years. Ghana is attempting to scale up its national health insurance scheme, which will significantly increase the proportion of individuals with green cards. This could push more patients from private to public health care, burdening public health institutions. Conversely, if private clinics like Ebenezer Clinic accept the national health insurance, discrimination may become a larger issue. Regardless, these are issues that need to be thought through as the government scales up the national health insurance plan.

Quality of care and cost form a duality of the most important factors affecting health decision making. Some individuals value quality over cost, while others care more about cost. Regardless of ranks, both play important roles in the process of choosing a health provider and seeking care. Distance from health providers may well serve as a further barrier, but its role must be further explored. Understanding the roles that these barriers play in the health decision making process of rural Ghanaians can inform policy makers and health providers ease the effects of these barriers and attract individuals to their health services.

Recommendations
Since Prampram Health Centre and Ebenezer Clinic are by the far the most popular in the area and are only about 200 meters apart, they are in a unique position to create an informal public-private partnership. Each facility does some things well- Prampram Health Centre has health promotion classes, while Ebenezer Clinic has better operating hours, prompt health workers, and a shorter waiting time. Each health facility can learn from the other to improve their health facility, attract more Ghanaians to utilize their services, and increase the health status of individuals in communities it serves.

Cost is a significant barrier to health access, especially for rural communities. A pro-poor health insurance scheme must be implemented. This is a recent goal that the government has yet to develop; about 30% of those interviewed had a government issued health insurance card. Research has shown that health insurance significantly increases access to health care. Policy makers must take into account the extra burden faced by individuals farther from health providers. Increased public transport and better roads are obvious structural improvements needed, but in addition a government financing system may want to consider travel vouchers for rural households. This would increase access and encourage rural communities to seek health care at orthodox health providers by reducing financial barriers.

To bolster the effects of a pro-poor health insurance scheme, the government must encourage health providers to promote health in rural communities with outreach programs. In Prampram sub-district, the only educational programs are maternity education programs that take place at Prampram Health Centre. Health centers should be encouraged to take health information to their constituents. Based on the time after illness that individuals seek health care, particularly at the Ebenezer Clinic, health promotion and disease prevention is an area where progress must be made in rural communities.

To decrease wait time at health providers, health workers should be prompt in arriving to work and seeing patients. Strict regulations should be implemented by the Ghana Health Service to ensure regular attendance. The government must do its part in establishing incentives for health workers; during fieldwork nurses refused to show up for work for a day after they had not been paid for the previous month. This is a tragic circumstance that negatively affects health worker attendance, health worker attitudes, and health outcomes for the ill.

The government should commit more effort into keeping skilled health workers in Ghana, as well as implementing periodic training for health workers. Many respondents list health worker attitudes as a major factor in determining health providers, so health providers must pay attention to the social aspects of health care Health workers. A greater focus on the training of the social aspects of care giving will further attract patients to orthodox health providers.
For other health providers not controlled by Ghana’s government, the results of this study should be seen as an opportunity to attract patients. By expanding services, increasing quality of drugs, maintaining stocks of government approved drugs, and improving hours of operation, health providers such as drug stores and drug peddlers can attract more business and provide better care.

This study invites further research on demand-side factors affecting health seeking behavior. The extent to which self-medication permeates the society, as well as the modes of self-medication that individuals use is an important step toward understanding what keeps individuals away from orthodox health providers.

More in depth cost analysis may also be researched to ensure that a government health insurance scheme is efficient and effective. To enhance the cost analysis, as well as the quality and seeking behavior factors, households can be examined by socio-economic status to determine to role of cost in health seeking behavior stratified by income level.

An analysis of communities close together, but with private and public health providers farther apart, may shed further light on the role of distance in health seeking behavior.

Conclusion
As Ghana develops, it will require the human capital of all of its citizens. In recent years, population growth has outstripped the provision of adequate health care, and access for rural Ghanaians has been adversely affected. In order to increase access Ghana’s government must focus on health provision from the supply side while taking into account the demand side. Increasing demand for health services and appeasing this demand with effective policies will significantly increase access and utilization of life improving health services.

Among demand side barriers, cost and perceived quality of health care providers are the most significant factors in the health decision making process. Individuals are unlikely to go to health providers because of high cost; instead they choose unregulated prayer camps and drug stores that are less successful than clinics in curing illness. This finding forces the government to look at the health financing system, from health insurance to road quality. Policy makers and health providers must take into account the additional hidden cost and opportunity costs such as transport costs and inability to accumulate income while at the health provider.

Quality of health care providers is another significant factor. This impacts the entire health system as they attempt to cure the sick with a caring touch. Nurses and other health workers need to be patient and understanding when explaining health instructions. Easing the burden of barriers to health care will empower patients to have choice of quality health providers and will foster greater utilization of health services. Recognizing and understanding demand-side barriers to health such as distance, cost, and quality of care will better inform policy makers in the quest for universal access and utilization of health services in Ghana.

References
1. Ghana Poverty Reduction Strategy Papers (PRSP) PowerPoint Presentation.

2. Van den Boom, G.J.M et al, “Health care provision and self medication in Ghana.” Institute of Statistical, Social and Economic Research, ISSER, University of Ghana, Legon.  March 2004.

3. Ensor, Tim and Cooper, Stephanie, “Overcoming Barriers to Health Service Access and Influencing the Demand Side Through Purchasing.”  Health, Nutrition and Population Discussion Paper, World Bank, September 2004.

4. Chikwama, Dr. Cornilius, “Inequality in Maternal Health: Applying Economic Methods to Guide Policy in Targeting the Poor.”  Immpact, 22 November 2007.  PowerPoint Presentation.

Appendix
Figure 1.