Service availability and utilization and treatment gap for schizophrenic disorders: a survey in 50 low- and middle-income countries
Antonio Lora a, Robert Kohn b, Itzhak Levav c, Ryan McBain d, Jodi Morris e & Shekhar Saxena e
a. Mental Health Department/Dipartimento di Salute Mentale, Lecco General Hospital/Ospedale di Lecco, via dell’Eremo 9/11, 23900 Lecco, Italy.
b. Department of Psychiatry and Human Behavior, Warren Alpert Medical School of Brown University and The Miriam Hospital, Providence, United States of America (USA).
c. Mental Health Services, Ministry of Health, Jerusalem, Israel.
d. Department of Global Health and Population, Harvard School of Public Health, Cambridge, USA.
e. Department of Mental Health and Substance Abuse, World Health Organization, Geneva, Switzerland.
Correspondence to Antonio Lora (e-mail: email@example.com).
(Submitted: 18 April 2011 – Revised version received: 15 September 2011 – Accepted: 21 September 2011 – Published online: 31 October 2011.)
Bulletin of the World Health Organization 2012;90:47-54B. doi: 10.2471/BLT.11.089284
Schizophrenic disorders are chronic and severe mental conditions that affect 26 million people worldwide and result in moderate or severe disability in 60% of cases.1 Due to their early onset and debilitating effects, schizophrenic disorders rank fifth among men and sixth among women as a leading cause of years lived with disability. Schizophrenic disorders also comprise roughly 1% of the global burden of disease (GBD), a fraction that is considered moderate to high. They also represent 1.3% of the disability-adjusted life years (DALYs) overall and 1,2%, 1.6% and 0.8% in upper-middle-income countries, lower-middle-income countries and low-income countries, respectively.2
Mental health services play a central role in the treatment of people with schizophrenic disorders, as they act both as direct providers of care and as supporters of primary care practitioners. Recent data indicate that in low- and middle-income countries, the treatment of people with schizophrenic disorders using first-generation antipsychotics and psychosocial interventions (family and psycho-educational), when delivered via a community-based service model, represents a cost-effective use of health resources.3,4 Despite this, only a minority of people with schizophrenic disorders receive care from formal mental health services.5
Access to specialized services is a key measure in evaluating the capacity of health-care systems to reduce the untreated burden of schizophrenic disorders. This measure requires information on how many people with schizophrenic disorders have access to care out of the total number of people in need of services. In addition to access, another key indicator is service utilization, which describes the services patients receive and the balance between outpatient and inpatient care. In an earlier report, Kohn et al.6 described the treatment gap as the absolute difference between the true prevalence of a disorder and the treated proportion of individuals affected by the disorder. The World Health Organization’s Assessment Instrument for Mental Health Systems (WHO-AIMS)7 comprises information on mental health systems in low- and middle-income countries and thereby allows, for the first time, an in-depth analysis of the availability of mental health services in these countries and a framework for ascertaining the accessibility of service delivery.
The goal of this paper is to utilize the WHO-AIMS instrument to analyse the accessibility of mental health services for people with schizophrenic disorders in 50 low- and middle-income countries, to estimate the magnitude of the treatment gap and to describe health service utilization among people affected with schizophrenic disorders.
The instrument: WHO-AIMS
The WHO-AIMS instrument consists of 155 input and process indicators covering six domains:7–8 (i) policy and legislative framework; (ii) mental health services; (iii) mental health in primary care; (iv) human resources; (v) public information and links with other sectors; and (vi) monitoring and research. This paper reports on selected indicators drawn from the second and forth domains.
Country-based investigators collected data for the indicators from all available sources (e.g. national and local statistics and surveys specifically planned to collect WHO-AIMS data) using clear definitions and explicit instructions. Country, regional and WHO headquarters-based staff provided technical assistance and supervision.
This paper, which focuses on mental health services for adults, draws on cross-sectional data from low- and middle-income countries or territories that completed a WHO-AIMS assessment between March 2005 and June 2010. Diagnostic data were available for only 50 of the 63 countries that completed this assessment. Thus, 13 countries were excluded because treated prevalence and treatment gap rates could not be calculated due to missing data and errors in data collection (e.g. contacts with users were reported instead of the number of users). However, the Mann–Whitney U test showed no significant differences between these 13 countries and the 50 countries used in the analysis of the nine selected indicators for which sufficient data were available (Table 1). The final sample included 46 countries, two local regions (a province of China and a state of India) and two territories (Kosovo and the West Bank and Gaza Strip). For convenience, all of them will be referred to as countries throughout this paper.
Table 1. Spearman’s correlation coefficients for associations between treatment gap, treated prevalence and facility utilization rates and preselected predictors obtained from the World Health Organization’s Assessment Instrument for Mental Health Systems
Of the 50 countries included in the sample, 15 were missing one of three data items: patients treated in outpatient facilities, community-based psychiatric inpatient units and mental hospitals. For these countries, we imputed the missing data item with the regional median value, weighted by population. Limited sample size precluded the inclusion of income group classification in this estimation. In total, we imputed nine data points for outpatients (Belize, Dominica, Georgia, Jamaica, Myanmar, Paraguay, Saint Lucia, Suriname and Tunisia) and six for inpatients in general hospital wards (Armenia, the Dominican Republic, Ecuador, Ethiopia, India (Gujarat) and the Maldives). On average, nine regional values were available to derive imputed values. However, in the case of Ethiopia, only four values were available, and in the cases of India (Gujarat) and the Maldives, only five. For this reason, imputations for these countries may be considered less reliable than for the rest.
Of the 50 countries, 11 were low-income, 30 were lower-middle-income and 9 were upper-middle-income countries according to World Bank criteria.9 The selected countries represented the following percentages of the populations of low- and middle-income countries situated in the six WHO regions: 11% in the European region; 13% in the Western Pacific region; 16% in the African region; 18% in the Region of the Americas; 26% in the South-East Asia region, and 57% in the Eastern Mediterranean region.
Treated prevalence and service utilization
Treated prevalence refers to the proportion of people with mental disorders served by mental health systems. The number of people per 100 000 population who received care for schizophrenic disorders in the various types of mental health facilities (outpatient facilities, community-based psychiatric inpatient units and mental hospitals) over the previous year can serve as a proxy for treated prevalence in specialized services. Population figures were based on United Nations 2004 estimates.10 The service utilization rate for each type of mental health facility (i.e. outpatient facility, psychiatric unit in a general hospital or mental hospital) was calculated as the number of people treated annually for schizophrenic disorders divided by the total number of patients with schizophrenic disorders treated on the whole in mental health facilities. While WHO-AIMS also provides information on patients treated in day treatment facilities and admitted to community residential facilities, these data were not incorporated, as diagnosis was not requested. However, overall the rates of utilization of these two types of facilities are modest compared with the rates of utilization of the facilities included in our estimates, and the absence of this information is unlikely to have substantially affected our estimates.
The treatment gap is the absolute difference between the true prevalence of a disorder and the proportion of affected individuals who are treated for the disorder. It is represented, in other words, by the proportion of individuals affected by schizophrenic disorders who fail to receive treatment. For each country, the treated prevalence of schizophrenic disorders (cases treated per 100 000 population) across all mental health facilities was compared with subregional prevalence estimates for schizophrenic disorders based on GDP data.2 In the GDP database, WHO Member States are grouped within each WHO region according to five mortality strata (denoted A to E), and this results in 14 subregions. Subregional estimates of the community prevalence of schizophrenic disorders range from a low of 270 per 100 000 in African subregion D to a high of 510 per 100 000 in European subregion B. The median treated prevalence for all subregions from which country data were derived was 430 per 100 000.
Initial descriptive analysis showed that the data were highly skewed in a positive direction. Accordingly, reported rates are median rates. For the multivariate analysis of predictors of access, treatment gap and service utilization, 52 WHO-AIMS indicators pertaining to organization, financing, resources and mental health service delivery were identified as possible predictors. However, complete data from all 50 countries was only available for 13 indicators, so only these indicators were used in the analysis. Spearman’s correlations were calculated to examine the relationship between each predictor (Table 1) and the five outcome variables (treated prevalence, treatment gap, rate of treatment in outpatient facilities, rate of inpatient treatment in mental hospitals and rate of inpatient treatment in psychiatric units in general hospitals). Given the degree of heterogeneity and variation in outcome scores, outlier analysis was not conducted. Backward step-wise regression analysis was used to determine the best-fitting predictive models for each of the outcomes. To be entered in the model, the independent variable could not be co-linear with the outcome measure nor part of the definition, and its correlation had to be significant at the P < 0.05 level based on the results of the Spearman’s correlations with the dependent variable. Variables whose correlation was significant at the P < 0.10 level were kept in the model.
Table 2 (available at: http://www.who.int/bulletin/volumes/90/1/12-089284) presents treated prevalence, service utilization rates and treatment gap for all countries in the sample. Treated prevalence in all specialized services was 128 per 100 000 population, with large variations by country income level: rates in upper-middle-income and lower-middle-income countries (126 and 157 per 100 000, respectively) were approximately four times higher than in lower-income countries (36 per 100 000).
Table 2. Descriptive statistics describing service utilization and level of treatment for each country, organized by World Bank income classifications
As shown in Table 1, treated prevalence was significantly associated with the estimated prevalence of schizophrenic disorders (r = 0.52), gross national income; availability of mental hospital beds (i.e. beds per 100 000 population); number of psychiatrists per 100 000 population and number of nurses in mental health facilities per 100 000 population. The best-fitting regression model explaining treated prevalence (Table 3) was the level of human resources – i.e. psychiatrists and nurses in mental health facilities – available to treat individuals with schizophrenic disorders.
Table 3. Summary statistics from backward step-wise regressions used to determine the best-fitting predictive model for each of the outcomes
The median value for treatment coverage was 31%. This suggests that roughly two thirds (69%) of the people with schizophrenic disorders were not receiving treatment. The treatment gap for schizophrenic disorders was larger in lower-income countries (89%) than in lower-middle-income (69%) and upper-middle-income countries (63%). The size of the treatment gap showed a significant negative association with the prevalence of schizophrenic disorders in the general population; gross national income; the availability of mental hospital beds; the number of psychiatrists per 100 000 population and the number of nurses in mental health facilities per 100 000 population. The best-fitting model accounting for the magnitude of the treatment gap included the numbers of psychiatrists and nurses in mental health facilities per 100 000 population.
Approximately 80% of patients with schizophrenic disorders were treated in outpatient facilities (Table 4). In terms of rates of schizophrenic disorders per 100 000 population, utilization rates for outpatient services were three times higher in upper-middle-income and lower-middle-income countries than in low-income countries. In upper-middle-income countries the rate of inpatient treatment in mental hospitals was six times higher than in low-income countries, and the rate of inpatient treatment in psychiatric wards in general hospitals was nine times higher.
Table 4. Median utilization ratesa for different types of facilities, by World Bank country income classification
Variables positively and significantly correlated with rate of treatment in outpatient facilities included the estimated prevalence of schizophrenic disorders; gross national income; the number of psychiatrists per 100 000 population, and the number of nurses in mental health facilities per 100 000 population. Similarly, variables showing a significant positive correlation with the rate of inpatient treatment in mental hospitals were the estimated prevalence of schizophrenic disorders; gross national income; the number of psychiatrists per 100 000 population, and the number of nurses in mental health facilities per 100 000 population.
The rate of inpatient treatment in psychiatric units within general hospitals was positively and significantly correlated with the percentage of mental hospitals organizationally integrated with outpatient mental health facilities, the number of beds in community-based psychiatric inpatient units and mental hospitals per 100 000 population, the percentage reduction in beds in mental hospitals during the previous five years, and the ratio of psychiatric beds located near large cities.
The best-fitting model explaining the outpatient treatment rate was the number of psychiatrists per 100 000 population and of nurses in mental health facilities per 100 000 population. For the rate of inpatients treatment in mental hospitals, only the number of nurses remained significant. Lastly, for the rate of inpatient treatment in psychiatric wards within general hospitals, the best-fitting model included only one predictor: organizational integrations of mental hospitals with outpatient mental health facilities.
The findings suggest that people with schizophrenic disorders in low- and middle-income countries have limited access to specialized mental health services. The median treated prevalence rate of 128 per 100 000 population per year is far lower than the figures suggested by community epidemiological studies (330 per 100 000 in Saha et al.11; 408 per 100 000 in the GBD, 2004 update).2
About two thirds of the people with schizophrenic disorders in low- and middle-income countries do not have access to specialized mental health care. The resulting treatment gap (69%) is much larger than the gap (32%) found by Kohn et al.,6 perhaps because different data sources were used in the two studies. Our study used data collected systematically by WHO-AIMS from mental health care providers in each country, whereas the earlier study used several community-based epidemiological surveys of individuals aged 15 and older that had been published since 1980 or provided by investigators or agencies. The differences between the two studies in the estimated population prevalence of schizophrenic disorders (from GBD estimates in this study and from epidemiological studies and reviews in the previous one) contribute only partially to the different results. We chose the GBD data because they provided subregional prevalence estimates. However, even if we applied the estimates produced by Saha et al., the treatment gap would remain substantially higher (62% in the whole sample and 86% in low-income countries) than in the report by Kohn et al. These differences can perhaps be partially explained by the fact that the sample of countries in the two studies was not the same: the Kohn et al. analysis included prevalence surveys from high-income countries, where the treatment gap is lower. In fact, differences between the two estimates of the treatment gap are reduced in a high-income country such as Italy.12 When Lora’s study in Italy was updated with the GBD 2004 prevalence estimates and was performed following the methods we used in this study (mental health service data collection and GBD estimates), the treatment gap in Italy (33%) was similar to the gap found by Kohn et al.6
These results do not account for differences in socioeconomic status and its effects on treatment gap or for regional inequities within a country. The gap is wider among those who have less (i.e. the poor, ethnic minorities, migrants) and need more (i.e. those among whom disorder rates are higher). For them accessibility is an issue and they require special programmes to bridge the disparities. Moreover, those who seek services are not always adequately treated or treated at all. We did not assess treatment adequacy. Hence, our data could be greatly overestimating the number of people who received appropriate treatment.
It is important to understand not only service accessibility, but also where people are receiving care. One of this study’s main findings is that the majority of people with schizophrenic disorders are treated in outpatient facilities, even in the most basic mental health systems of many low-income countries. Outpatient care effectively increases coverage within a mental health system. Inpatient mental health facilities, whether in general hospitals or mental hospitals, only modestly contribute to overall service accessibility.13
Our data clearly show that specialized mental health services alone are unable to cope with the burden of schizophrenic disorders in low- and middle-income countries. From a public health perspective, primary care services should fill this gap by delivering effective packages of care in collaboration with specialized services. WHO has recently launched the WHO Mental Health Global Action Programme and made a case for an integrated approach that emphasizes the role of the primary care sector in scaling up care.4
In terms of predictors of service utilization, the level of available human resources, in terms of psychiatrists and nurses in mental health facilities, appears to positively predict treatment prevalence and rate of outpatient care and negatively predicts the overall treatment gap. These results confirm the need to scale up the workforce in low- and middle-income countries.14
Our findings with respect to the rate of inpatients treated in general hospitals support the fact that the higher the availability of psychiatric beds in community medical facilities, the higher the rate of treatment in the psychiatric units of general hospitals. Unfortunately, the rate of utilization of general hospitals remains low in low- and middle-income countries.
This study has limitations which stem primarily from the scarcity of reliable databases in low- and middle-income countries. The reliability and validity of the information reported by these countries are therefore questionable and our estimate of treated prevalence may be biased and either too low or too high. On the one hand, WHO-AIMS does not request that data from day-treatment facilities and community residential facilities be broken down by diagnosis. For this reason, patients with schizophrenic disorders who were attended in these facilities were not counted, and this could have resulted in an underestimation of treated prevalence. However, results from the WHO-AIMS show that in low- and middle-income countries these types of facilities are rare and contribute only 1% of the overall treated prevalence.13 In addition, a few country-based investigators had difficulty obtaining information from private mental health facilities and nongovernmental organizations (NGOs) involved in mental health care. However, in low- and middle-income countries access to private mental health facilities is primarily limited to the wealthy; few NGOs treat people with serious mental disorders. Thus, the additional coverage provided by the private sector and NGOs would be small. On the other hand, the fact that some patients could have been treated in more than one setting (e.g. in both a community-based inpatient unit and an outpatient clinic within the same year) and been counted more than once could have resulted in an overestimation of treated prevalence.
The diagnoses provided in the WHO-AIMS, which are based on administrative data, may be of poor quality. For example, the huge rates of treated prevalence for schizophrenic disorders found in Latvia and the Ukraine, possibly resulting from misdiagnosis, made us cap the treated prevalence estimates for those countries at 100%. However, grouping of diagnoses from the International classification of diseases, tenth revision into large diagnostic classes, as has been done in the WHO-AIMS, may increase their validity. This is because differentiating between classes of disorders (e.g. schizophrenic disorders versus affective disorders) is perhaps easier than differentiating within classes of disorders (e.g. schizophrenic disorders versus schizoaffective disorders).
Lastly, the countries included cannot be assumed to be representative of their regional areas. The sample of 50 countries is not large enough and our data represent a median of 17% of the population of each country’s respective region. Moreover, since 30 out of the 50 countries included in this report are in the lower-middle-income category, the overall findings are largely reflective of countries in this income group.
Service availability, service utilization and treatment gap constitute key indicators for evaluating the capacity of mental health systems to respond to the needs of people with schizophrenic disorders. Improvement of mental health systems, particularly service accessibility, could be monitored through these indicators. Due to uncertainty in community prevalence estimates, the treatment gap analysis is an approximation of the need for care; however, such measurements are necessary to monitor coverage at the mental health system level. From this perspective, these indicators may become important in conducting advocacy, and the information they provide could raise awareness among governments and stakeholders aiming to plan and deliver better mental health care.15
- Eaton WW, Martins S, Nestadt G, Bienvenu J, Clarke D, Alexandre P. The burden of mental disorders. Epidemiol Rev 2008; 30: 1-14 doi: 10.1093/epirev/mxn011 pmid: 18806255.
- The global burden of disease: 2004 update. Geneva: World Health Organization; 2008. Available from: http://www.who.int/healthinfo/global_burden_disease/estimates_regional/en/index.html).
- Chisholm D, Gureje O, Saldivia S, Calderón M, Wickremasinghe R, Mendis M, et al., et al. Schizophrenia treatment in the developing world: an interregional and multinational cost-effectiveness analysis. Bull World Health Organ 2008; 86: 542-51 doi: 10.2471/BLT.07.045377 pmid: 18670667.
- mhGAP intervention guide for mental, neurological and substance use disorders in non-specialized health settings: mental health Gap Action Programme (mhGAP). Geneva: World Health Organization; 2010.
- Saxena S, Lora A, Morris J, Berrino AM, Esparza P, Barrett T. Mental health services in 42 low and middle income countries: a WHO-AIMS cross-national analysis. Psychiatr Serv 2011; 62: 123-5 doi: 10.1176/appi.ps.62.2.123 pmid: 21285088.
- Kohn R, Saxena S, Levav I, Saraceno B. The treatment gap in mental health care. Bull World Health Organ 2004; 82: 858-66 pmid: 15640922.
- World Health Organization Assessment Instrument for Mental Health Systems (WHO-AIMS). Geneva: World Health Organization; 2005. Available from: http://www.who.int/mental_health/evidence/WHO-AIMS/en/index.html [accessed 20 October 2011].
- Saxena S, Lora A, van Ommeren M, Barrett T, Morris J, Saraceno B. WHO's Assessment Instrument for Mental Health Systems: collecting essential information for policy and service delivery. Psychiatr Serv 2007; 58: 816-21 doi: 10.1176/appi.ps.58.6.816 pmid: 17535942.
- Country classifications [Internet]. Washington: World Bank; 2009. Available from: http://data.worldbank.org/about/country-classifications [accessed 20 October 2011].
- World population prospects, 2004 revision. New York: United Nations Population Division; 2005.
- Saha S, Chant D, Welham J, McGrath J. A systematic review of the prevalence of schizophrenia. PLoS Med 2005; 2: e141- doi: 10.1371/journal.pmed.0020141 pmid: 15916472.
- Lora A. An overview of the mental health system in Italy. Ann Ist Super Sanita 2009; 45: 5-16 pmid: 19567972.
- Mental health systems in selected low and middle-income countries: a WHO-AIMS cross-national analysis. Geneva: World Health Organization; 2009.
- Bruckner TA, Scheffler RM, Shen G, Yoon J, Chisholm D, Morris J, et al., et al. Scaling up the mental health workforce in low and middle income countries. Bull World Health Organ pmid: 21379414.
- Saraceno B, Freeman M, Funk M. Public mental health. In: Detels R, Beaglehole R, Lansang MA, Gulliford M, editors. Oxford textbook of public health. 5th ed. Oxford: Oxford University Press; 2009.