Access to mental health care in El Salvador: a case study of progress toward decentralization.

OBJECTIVE: To assess the need for decentralization of psychiatric services in El Salvador, based on country-specific evidence, and to generate baseline measures the government and researchers could use to monitor and measure future progress toward decentralization.
METHODS: Observations were made and psychiatric outpatients and their families/caregivers (n = 453) surveyed with a cross-sectional instrument at the National Psychiatric Hospital (Hospital Nacional General y de Psiquiatría "Dr. José Molina Martínez") in the capital city of San Salvador. Findings were compared with national census data to determine barriers to psychiatric care in a largely centralized system. The following data were collected: department of origin, travel time to hospital, time in hospital, costs, and qualitative comments about accessing services. Descriptive statistics and regression were used to assess the data. Government reports on all psychiatric outpatient consultations provided nationwide in 2015 (n = 61 010) at 10 regional hospitals were also reviewed and compared to population data to determine the proportion of the population of each hospital service area (department or group of departments) that received a psychiatric outpatient consultation.
RESULTS: Households from the half of the country (7 out of 14 departments) farthest away from the National Psychiatric Hospital (42% of the general population) spent twice as much time and three times as much money to access the hospital's psychiatric outpatient services, resulting in severe hardship on both patients and families/caregivers.
CONCLUSIONS: This report estimates that 45% of those with mental health needs in the seven departments farthest away from the National Psychiatric Hospital departments are still not accessing services compared to the seven departments nearest the hospital. The results of this study support the World Health Organization's call for governments to fully implement community-based mental health systems. This is the first study to assess progress toward decentralization of psychiatric services in El Salvador.

Mental and substance use disorders are the leading cause of years lived with disability (YLDs), with youth ages 10–29 years suffering the highest proportion of disability-adjusted life years (DALYs) (1). In Latin America and the Caribbean (LAC), mental disorders represent 24.5% of the burden of disease, far more than any other group of illnesses (2), yet the majority of governments in Central America spend only 1% of their national health budgets on mental health. In El Salvador, 92% of the budget for mental health goes to cover the costs of the National Psychiatric Hospital (Hospital Nacional General y de Psiquiatría “Dr. José Molina Martínez”) (3).

Studies have demonstrated the importance and benefits of decentralization of mental health services. These include patients receiving outpatient care at a rate three times higher than in areas with centralized services (4), reduced suicide rates (5), lower costs (6), and reduced use of jails as holding cells for people with mental conditions (7). In India, a study of decentralization showed that nearly 80% of patients with mental illness for a year or longer coming to a new urban treatment facility had no past psychiatric treatment or only irregular treatment. Two of the six reasons given by patients for not seeking treatment previously were distance and cost (8).

Since the Declaration of Caracas in 1990, advocates have been calling for the decentralization of mental health services across the LAC region (9). However, continued, recent calls indicate that progress is limited (10–13).

The pace of change to community-based mental health systems has been very slow in LAC countries, with the exception of Brazil, Chile, and Panama, as well as Belize (14–17). All four countries reduced the number of mental hospital beds by a significant amount (62% over a 10-year period in Brazil, Chile, and Panama, and 55% in Belize over a similar time period), in contrast to other countries in the region, who reduced the number of beds by only 24%. This reform was facilitated in the first three countries by increasing the percentage of public mental health expenditures allocated to general hospitals, outpatient facilities, and community services—68% in Brazil, 88% in Chile, and 56% in Panama; Belize opted to decentralize its mental health care program by training more primary care psychiatric nurses in each health district (14).

In El Salvador, some progress has been made. For example, by the end of 2015, psychiatrists had been placed in 11 out of 30 of the country's public hospitals, and primary care physicians were being trained to implement the World Health Organization (WHO) Mental Health Gap Action Programme (mhGAP), which provides a set of manuals on mental health treatment and referral protocols to help doctors provide mental health treatment at the community level. Yet the government's commitment to increasing access to mental health services remains unclear, given that as of December 2016 the Ministry of Health (MoH) had trained only 365 or 5.7% of public sector doctors in the mhGAP protocols, and the program has come to a standstill. To the best of the authors’ knowledge no previous assessment of decentralization of mental health services in El Salvador has been published.

The objective of this study was to assess the need for decentralization of psychiatric services in El Salvador, based on country-specific evidence, and to generate baseline measures the government and researchers could use to monitor and measure future progress toward decentralization.

MATERIALS AND METHODS

Data collection for this case study consisted of three parts: 1) a cross-sectional survey of 453 psychiatric patients and their family members/caregivers seeking mental health care services at the outpatient clinic of the National Psychiatric Hospital in San Salvador, including both quantitative and qualitative questions; 2) observations of care-seeking by the same sample group; and 3) a review of MoH reports on outpatient psychiatric consultations at 10 regional hospitals (18) and comparison with population census data. This approach was used to triangulate findings and to collect a broader range of information, including qualitative data that would help explain the impact on families of the time and distance data collected in the quantitative questions.

Survey subjects were interviewed over a period of three months, from November 2015 to January 2016. Participants came from all 14 departments of El Salvador. Data were collected at the household level, using a brief-screening method (surveys of 2–3 minutes), so no personal or demographic information was collected. The researchers talked with as many households as possible, at the clinic. Household members included users of mental health services, family caregivers, and any other person(s) who had a primary and long-term role in supporting the person with the mental condition. Persons who were unwilling or unable to respond to the questions were excluded from the study and totaled less than 5% of those screened.

During the three months of the study, on the majority of days the clinic was open, the team of one to three researchers arrived just prior to the 6:30 am, 10:30 am, and 12:30 pm time slots to speak with people waiting for their appointments. An announcement was made in the waiting room describing the study, and the survey; the researchers, and their professional affiliations; and how the researchers would conduct the interviews. The researchers spoke to each household privately, beginning the interview by asking if they agreed to answer a few questions, and letting them know that they could refuse, without repercussions, and that their participation was anonymous. Due to contextual factors and to ensure anonymity of subjects, an oral approach, rather than signed informed consent, was used.

The survey consisted of a few structured and semi-structured questions. Some of the households (40) provided additional commentary about their experiences, voluntarily. The methodology for assessing these qualitative data is described below. Variables of interest included department of origin; travel time to hospital; time in hospital; total cost and breakdown of costs (transportation, food, lodging); type of transportation (foot, bus, taxi); number of people who came together (household group); and qualitative comments about experiences accessing care. Descriptive statistics and regression were used to analyze/summarize the quantitative data. Distance was the predictor variable, while time, cost, and number of households by department were the outcome variables.

For the qualitative analysis, responses to questions as well as unsolicited comments were recorded for each of the 163 households screened. Using simple content analysis, comments were assessed by hand by the first two authors (SVN and MCT) and categorized by consensus into common themes (listed, with sample comments for each one, in Table 1). The authors then reviewed the final themes and representative comments for face validity and comparison with the conclusions drawn from the quantitative data. A group of eight psychiatrists from the National Psychiatric Hospital later reviewed the findings and had no disagreements with either the qualitative or quantitative data and findings.

TABLE 1

Sample illustrative comments from households (patients/families/caregivers) surveyed in mental health case study, by theme, about challenges in accessing care at the National Psychiatric Hospital, San Salvador, El Salvador, November 2015–January 2016

Theme (%a)	Sample illustrative comments
Human rights abuse (5%)	“When I'm at the hospital I have to leave my loved one locked in the house.” “While I go to the appointment for the day, I have to leave him tied up.”
Physical hardship (5%)	“I spent the night in the hospital, sleeping on the ground.” “I spent the night sleeping here in this seat.”
Financial burden—hunger (30%)	“I didn't eat because I don't have money. I'll stand it until I get back to the house.” “We came in a taxi, but now there's not enough money left for us to eat.”
Financial burden—transportation (20%)	“We paid $70 to come in a taxi from Usulután because our ill loved one cannot handle the trip on a public bus.” “We pay a neighbor to take us in his car. We pay $120 to him, and food is about $20 for all of us.”
Financial burden—no bus money (7.5%)	“I came on foot because I don't have bus fare (25 cents).” “I walked. I have no money for the bus.”
Financial burden— spent money on lodging (32.5%)	“We come from the interior. We have to spend the night in San Salvador in a guest house.” “There isn't time to get here for our morning appointment, so we have to come the day before.”

Proportion of all comments from the 25% of the total survey sample (40 of 163 households) that provided feedback about their experience seeking care. As the study did not explicitly ask everyone in the sample for comments, the authors believe that far more than 25% of care-seekers experience these challenges. Further, it is unknown how many are not coming to the hospital as a result of these barriers.

Data on the number of psychiatric outpatient visits available at all 30 hospitals in the country were collected from MoH reports available on the ministry website (18). Only 10 of the 30 hospitals reported psychiatric consultations for 2015 (n = 61 010, with more than half reported by the National Psychiatric Hospital in San Salvador). The research team, with support from consultants, developed calculations to compare the number of consultations by department with population census data by department to estimate the percentage of persons by department that should be receiving mental health services if there were an even distribution of services across the country.

The study was carried out in accordance with ethical protocols of human subject research and approved by both the institutional review board at James Madison University (protocol #15-0463) and the National Psychiatric Hospital ethics committee.

RESULTS

As the psychiatric patients and their families/caregivers were highly unreliable in identifying the current diagnosis, hospital statistics were used as a proxy for estimating sample proportions of illness type. For 2015, the top diagnoses treated by psychiatrists in the outpatient clinic of the National Psychiatric Hospital were epilepsy (5 287 persons), depressive episode (3 590), organic mental disorder (3 525), paranoid schizophrenia (2 745), organic affective disorder (1 959), and mixed anxiety and depression (1 546). Men were seen for organic mental disorder and schizophrenia more than women, while women were seen far more frequently for mood disorder complaints. Treatment consisted of minimal and irregular psychoeducation by a nurse, followed by a meeting with a psychiatrist for the patient and family member/caregiver and provision of medication prescriptions. A social worker or psychologist was sometimes available to discuss various topics with patients and families, including substance abuse and other issues.

Travel time and cost

It took families an average of more than 2 hours to travel to the hospital (mean: 2.15 hours, margin of error of 6 minutes; median: 2.00 hours; standard deviation (SD): 2.01 hours, range: 0.08–12.00 hours; n = 448), or more than 4 hours roundtrip (Figure 1). Those who lived nearby walked 5 or 10 minutes, but others traveled for 4 hours or more. Doubling the average one-way 2-hour figure, the roundtrip was more than 4 hours. When time waiting in the hospital is included, the average trip was more than 8 hours. As a result, families often lose a day of income for each trip. Some families spent the night near the hospital so they could catch an early bus to the hospital. A few slept in chairs or on the hospital floor for the night before their early morning appointments.

FIGURE 1

Time required (hours) for psychiatric patients/families/caretakers to travel to outpatient consultations at the National Psychiatric Hospital, San Salvador, El Salvador, November 2015–January 2016

Caregivers who live far away could not always bring their loved one who was ill, usually because the patient had physical or emotional limitations that prevented him/her using public transportation over long distances, and the family could not afford to take a taxi such a distance. Some family caregivers commented that while they went to the hospital to keep the medical appointment, they had to leave their patient tied up or locked in the house.

Household expenses were primarily related to transportation, meals, and sometimes overnight accommodation. The average expenses per household were just under US$ 13, but some households spent much more, with the four households that spent the most spending between US$ 100 and US$ 210 (mean: US$ 12.57, margin of error US$ 0.62; median: US$ 9.00; SD: US$ 17.36, range: US$ 0–$210; n = 450) (Figure 2). These amounts are significant compared to the minimum wage in the country, which averages US$ 5.33 per day.

FIGURE 2

Money spent (US$) by psychiatric patients/families/caretakers to travel to outpatient consultations at the National Psychiatric Hospital, San Salvador, El Salvador, November 2015–January 2016

El Salvador uses U.S. dollars as its currency.

This study did not ask or include information about purchases of medications from private pharmacies when the prescribed medications were not available free from the hospital. The cost of medications in El Salvador is a complex issue beyond the scope of this report. For example, while a psychiatrist may be available in a rural regional hospital for a consultation, medications must often be obtained from the National Psychiatric Hospital, meaning an additional trip.

Time in the hospital

The mean time for being at the hospital is almost 4 hours (mean: 3.73 hours, margin of error 17 minutes; SD: 1.27, range: 1–9 hours; n = 163) (Figure 3). Based on observations and interviews, it is clear that families have to wait in several long lines in the outpatient clinic, including 1) the registration line, to sign in; 2) the line for the psychiatrists; and 3) the pharmacy line. The authors observed 80 or more people in the pharmacy line on multiple occasions.

FIGURE 3

Time spent (hours) by psychiatric patients/families/caretakers for outpatient consultations at the National Psychiatric Hospital, San Salvador, El Salvador, November 2015–January 2016

Does not include time spent at hospital overnight waiting for morning outpatient appointment.

In 2016, the authors of this report surveyed the National Psychiatric Hospital pharmacy and seven public regional hospital pharmacies by phone to determine whether the medications that doctors and psychiatrists are allowed to prescribe were available. Results indicated that second-generation psychiatric medications were not often available; first-generation medications were available more frequently, but no medications were consistently available all year long. On average, according to hospital pharmacists, requested medications were available about half the time. If the hospital pharmacy does not have a medication, the family is told to call back regularly to see if it is available. However, hospital pharmacies sometimes do not answer the phone, despite repeated calls. As a result, families are often forced to make repeated trips to hospital pharmacies in search of needed medications.

Number of people by department

El Salvador has 14 departments comprising a population of 6.3 million. The largest percentage of households in the sample for this study (34.6%) came from the department of San Salvador, where the National Psychiatric Hospital is located. The four departments around San Salvador had the next largest number of households seeking care (about 10% each). The smallest number of people came from the departments farthest from San Salvador—to the East, Morazán (1.3%), and to the West, Ahuachapán (1.6%).

Comparison of time and cost by department

Travel time and expenses for families were compared by two sets of department of origin: the seven departments nearest to the National Psychiatric Hospital and the seven farthest away from it. Families from the second group had to travel twice the time to access outpatient services compared to families from the first group (3.76 hours versus 1.85 hours each way). Families from the second group also spent almost three times more than the first group (US$ 27.48 versus US$ 9.83) (Table 2). Some people only spent a little, even though they traveled quite a distance, because after paying money for the buses they had no money left for food or lodging. These patients and families ate little to nothing at all while seeking/waiting for care. For them, it is a great sacrifice to access treatment.

TABLE 2

Proportion (%) of all psychiatric outpatient consultations provided nationwide compared with proportion (%) of population, by groups of departments/states served by 10 regional hospitals,a El Salvador, 2015

Regional groups of departments/states	Share of all psychiatric outpatient consultations provided nationwide (n = 61 010)		Populationb	Share (%) of national population	Share (%) of psychiatric outpatient consultations versus share (%) of population (% +/–)c	Proportion of population that received a psychiatric outpatient consultation
	No.	%				%
1. San Salvador, Cuscatlán, La Paz, Cabañas, San Vicente	37 264	61.1	2 661 102	42.3	+44.4	1.40
2. Chalatenango	1 602	2.6	204 808	3.3	–21.2	0.78
3. La Libertad	4 482	7.4	747 662	11.9	–37.8	0.60
4. Sonsonated	3 840	6.3	463 739	7.4	–14.9	0.83
5. Santa Ana, Ahuachapánd	6 513	10.7	905 487	14.4	–25.7	0.72
6. Usulutánd	2 971	4.9	366 040	5.8	–15.5	0.81
7. San Miguel, La Unión, Morazánd	4 338	7.1	941 582	15	–52.7	0.46
Total	61 010	100.1	6 290 420	100.1

Group 1 (in and around the capital city area) has 4 hospitals, including the National Psychiatric Hospital (Hospital Nacional General y de Psiquiatría “Dr. José Molina Martínez”), while Groups 2–7 have one hospital each.

Based on 2013 population survey (Encuesta de Hogares de Propósitos Múltiples 2013, Gobierno de la República de El Salvador, Ministerio de Economía, Dirección General de Estadística y Censos (http://www.digestyc.gob.sv/EHPM2013/digestyc/resultado.pdf).

Proportion of total psychiatric outpatient consultations (%) minus proportion of national population (%) divided by proportion of national population. For example, for San Salvador, Cuscatlán, La Paz, Cabañas, and San Vicente: 61.1–42.3 = 18.8 / 42.3 = 44.4%. Region 1 is thus providing 44.4% more consultations than one would expect based on their population compared to other regions.

These seven departments are farthest away from the National Psychiatric Hospital.

To see if there was a significant correlation between distance (the independent variable) and 1) time, 2) cost, and 3) number of families, by department, who came to the National Psychiatric Hospital for outpatient services, regression analysis was used. The analysis showed strong effect sizes for all three pairs of variables: distance and time to arrive at the hospital (P < 0.001; R2: 0.29; R: 0.54; F(1,446): 180.341); distance and costs (P < 0.001; R2: 0.28; R: 0.53; F(1,448): 176.110); and distance and number of families, by department, who came to the National Psychiatric Hospital for outpatient services (P = 0.022; R2: 0.36; R: 0.60; F(1,12): 6.880).

Table 1 lists qualitative themes and sample comments that reflect the impact of time and cost on family burden. The most common theme was the burden of spending money on lodging, mentioned by 32.5% of the survey participants who made unsolicited comments (40 out of 163 households, or 25%). This was followed by the burden of not having enough money for both transportation and food, leaving the survey participant(s) hungry (30%). Next was the burden of spending on transportation (20%), followed by those who had no money for even public bus rides and so had to walk to the hospital (7.5%), then physical hardship (5%), and resulting human rights abuses (5%).

Analysis of MoH data on psychiatric outpatient visits

Because the survey data did not include information from other hospitals where patients were receiving psychiatric outpatient services, and to take into account that the MoH had begun to decentralize some psychiatric outpatient consultations, the authors reviewed documents on the MoH website to obtain the number of psychiatric outpatient consultations in 2015 provided at all 30 hospitals in El Salvador and compared these data with census data by department. A total of 10 hospitals reported psychiatric outpatient consultation data, 60% of which came from the seven departments nearest to the National Psychiatric Hospital.

In the seven departments nearest to the National Psychiatric Hospital, five general hospitals reported offering psychiatric outpatient consultations. Two of these hospitals were in the capital city of San Salvador, relatively close to the National Psychiatric Hospital. In the seven departments farthest from the National Psychiatric Hospital, four hospitals reported offering psychiatric outpatient consultations (Table 2). The second-to-last column of Table 2 shows what percentage each department (or group of departments that share the same hospital(s) where psychiatric consultations are provided) reported above or below what would be expected if services were decentralized based on the respective proportion of the general population. The department group that included San Salvador was the only one that reported a greater number of consultations (44.4% higher) than what would be expected based on its population. All other departments reported a lower proportion of consultations (–14.9% to −52.7%) than would be expected based on their share of the total population. The calculation used was a simple equation based on the population and share of consultations compared across departments.

Using information from the MoH reports, and population data, the authors estimated that 45% of the population in the seven departments farthest away from the National Psychiatric Hospital lack access to psychiatric services, despite efforts to decentralize consultations at the end of 2015. This translates into 14 450 families without access.

DISCUSSION

This study of decentralization of mental health services in El Salvador supports the concerns of LAC mental health leaders and researchers that progress toward community mental health services has been slow and is far from being achieved. Through both quantitative and qualitative data, the study adds evidence to the research demonstrating that there are significant time, cost, and emotional burdens on households when they have to travel to centralized services for mental health care. For example, this study estimates that 45% of people in need of mental health services are not accessing those services in the half of the country farthest from the National Psychiatric Hospital, possibly due to barriers of distance, time, and cost.

The qualitative comments indicated that people suffer a great deal from the challenges and obstacles inherent in a mental health system organized around a centralized psychiatric hospital for an entire country. Families have to make difficult choices, selecting a strategy from multiple bad options to obtain treatment, such as giving up daily income to attend the appointment. In El Salvador it is often a family caregiver or friend who goes to the appointment because the person with mental health problems often can not endure the travel due to physical, age, or emotional limitations, or the family simply does not have enough money. Families and users make difficult choices to travel long distances for treatment, including going without food or sleeping in chairs overnight at the hospital.

This report provides baseline measures that government and researchers could use to assess future progress toward decentralization of mental health services. These include the number of outpatient consultations provided outside the National Psychiatric Hospital, the number of hospitals providing psychiatric outpatient consultations, the number of health personnel trained in mhGAP, the distance traveled and costs incurred by households to access treatment, waiting times, and qualitative impacts.

To the best of the authors’ knowledge, this type of study has not been carried out before in Central America. While recognizing that analysis of census and health care use (“consultation”) data has limitations, this methodology is easy to implement and appears to be effective in determining a potential percentage of the population without access to mental health services. Although the data came from self-reports, the sample was sufficiently large, the response distribution curves of the raw data were normal, and the face value and consistency of responses was high. Taken together these indicate a high level of validity.

Generalizability of results is limited because the survey portion of the study sampled only one hospital. Still, the socioeconomic conditions found in El Salvador are similar to many other low- and middle-income countries, so the authors believe the conclusions are likely applicable to similar countries. In addition, quantitative studies from Norway (4), Finland (5), Spain (6), Canada (7), India (8), and Kenya (19) support the conclusions provided here. Further studies at the country level could strengthen the urgent call by WHO for decentralization of mental health services.

Limitations

Due to the brief-screening nature of the data collection method used in this survey, no demographic, medical, or household composition data were collected, which limited the scope of the study results. Other limitations included 1) missing data, which led to different sample sizes for different survey questions, and 2) the lack of good-quality epidemiologic data by department, which resulted in the authors’ reliance on census and consultation data.

Conclusions

This study of decentralization of mental health services in El Salvador adds evidence to the broader literature on LAC countries that expresses concern about the slow and uneven progress toward community-based services in the region. A large percentage of households in the seven departments farthest from the National Psychiatric Hospital in El Salvador appear to be unable to access services. Burdens are significant and onerous for households to access to care. The authors offer baseline measures that government and researchers could use to assess future progress toward decentralization of mental health services.

Recommendations

The authors recommend that the Ministry of Health in El Salvador 1) use the data reviewed in this study as baseline measures to help monitor progress toward decentralization of mental health services, and 2) urgently implement a full decentralization process to reduce the burden of time and cost for those seeking care and to increase access to medications. Other recommendations include more extensive and rigorous research, including, 1) a national-level mental health epidemiology study to generate the very important yet still undetermined “needs” measure required to capture the impact of barriers such as stigma, denial, noncompliance, long waiting lines, and poverty that are not addressed in the census and consultation data used in this study, and 2) a more rigorous study capturing the influences of distance, time, and cost of travel to all main hospitals in the country that provide mental health services, as well as other barriers to care, to assess and strengthen this study's findings, which could further support the call for decentralization.