The Eastern Mediterranean Region (EMR) represents twenty-two countries. These are Afghanistan, Bahrain, Djibouti, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Pakistan, Palestine, Qatar, Saudi Arabia, Somalia, Sudan, Syria, Tunisia, United Arabs Emirates, and Yemen (1).
World Health Organization Eastern Mediterranean Countries (WHO EMC) spent US$ 92 billion on health in 2008. This accounts for 1.6 percent of global health spending for 8 percent of the world's population. The six high-income countries in the region spent approximately 34 percent of this amount, representing 6.6 percent of the region's population. In the area, 7.8 percent of general government expenditure is allocated to health. The per-capita total expenditure on health is USD 326 for WHO EMC (Reference Boncz, Kaló, Mohamed Ibrahim and Greenberg2). Furthermore, out-of-pocket payment accounts for approximately 60 percent of health expenditure in most low-income countries and reaches 80 percent in some countries. Healthcare insurance plans help to limit the financial burden on families and ensure proper access to health services (3). Thus, it is important to make efficient use of the limited resources to reduce the burden on poor populations (Reference Desai, Chandwani and Rascati4).
Results of pharmacoeconomic studies are of great importance to WHO EMC healthcare decision makers and can help in managing healthcare costs efficiently in their countries. Many Western countries have developed guidelines for conducting economic analyses (Reference Hjelmgren, Berggren and Andersson5–8). Health economics data have been evaluated in many countries; however, these data are limited in most developing countries (9;Reference Lee, Brouwer, Lee and Koo10) where the available data may be of poor quality (Reference Gavaza, Rascati, Lawson, Lawson and Mann11–Reference Machado, Iskedjian and Einarson15). The quality of health economic data is important for medical decision making, and poor-quality data may not be useful. Health economic publications from the WHO EMC are limited (Reference Boncz, Kaló, Mohamed Ibrahim and Greenberg2).
Several studies have evaluated the quality of health economic publications in India (Reference Desai, Chandwani and Rascati4), Korea (Reference Lee, Brouwer, Lee and Koo10), Zimbabwe (Reference Gavaza, Rascati, Lawson, Lawson and Mann11), Nigeria (Reference Gavaza, Rascati, Oladapo and Khoza12), Bangladesh (Reference Hoque, Khan and Hossain13), South Africa (Reference Gavaza, Rascati, Oladapo and Khoza14), South America (Reference Machado, Iskedjian and Einarson15), Thailand (Reference Teerawattananon, Russel and Mugford16), and Russia (Reference Gavaza, Shepherd, Shcherbakova and Khoza17). However, little is known about the quality and the extent of reporting recommended information in health economic studies in WHO EMC. To our knowledge, no published study has systematically reviewed pharmacoeconomic studies conducted in WHO EMC. Thus, the objective of this study was to evaluate the extent of reporting recommended information in published health economic research in WHO EMC.
METHODS
The literature was searched in March 2014 using PubMed, MEDLINE and Google Scholar to identify pharmacoeconomic studies in WHO EMC. Search terms included “pharmacoeconomic,” “health economic,” “cost,” “cost-effectiveness analysis,” “cost-minimization analysis,” “cost-utility analysis,” “cost-benefit analysis,” “economics,” “pharmacy,” “pharmaceuticals,” and “Afghanistan,” “Bahrain,” “Djibouti,” “Egypt,” “Iran,” “Iraq,” “Jordan,” “Kuwait,” “Lebanon,” “Libya,” “Morocco,” “Oman,” “Pakistan,” “Palestinian territories,” “Qatar,” “Saudi Arabia,” “Somalia,” “Sudan,” “Syrian Arab Republic,” “Tunisia,” “United Arab Emirates,” and “Yemen.” These keywords were used alone and in different combinations for each country.
The inclusion criteria were as follows: (i) original studies; (ii) comparisons between pharmacy services, treatment modality, or against no treatment; (iii) conducted in WHO EMC; and (iv) manuscript published in English. The exclusion criteria were as follows: (i) being multiple-country comparisons, (ii) having no cost analyses, (iii) being only a cost-of-illness study. Two researchers carried out the literature search and identified articles independently. All articles that met the inclusion criteria were confirmed by a third researcher.
Data collection was done using the data collection form that was developed by Gavaza et al. (Reference Gavaza, Rascati, Lawson, Lawson and Mann11) and has been used in previous studies (Reference Desai, Chandwani and Rascati4;Reference Gavaza, Rascati, Oladapo and Khoza12;Reference Gavaza, Shepherd, Shcherbakova and Khoza17;Reference Jiang, Ma, Desai, Yang and Rascati18). The form has two main parts (i.e., general information and economic analysis information). The first part included items on the number of authors, country of residence of primary author, training of primary author, year of publication, journal in which the study was published, and the country of the journal. The second part included items on method of economic evaluation, types of costs included, the primary outcomes, perspective, study design, and funding source.
The Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist, which was developed for better reporting of economic evaluations, and ultimately, better health decisions, was used to assess the level of completeness of full pharmacoeconomic studies. The checklist includes twenty-four items and accompanying recommendations, which are contained in the user friendly checklist (Reference Husereau, Drummond and Petrou19). Three scoring points for each item was used-full score = 1, a midpoint score = 0.5, or a zero. Finally, each economic evaluation had a combined score out of 24 (acceptable level of completeness, ≥ 20; fair level of completeness, 14–19; not acceptable level of completeness, ≤ 14).
Two blinded reviewers assessed the abstracts of the identified studies independently. All abstracts that met the inclusion criteria were confirmed by a third researcher. Furthermore, disagreements between the two reviewers were resolved through discussions and assessment by a third reviewer. If the difference between the scores given by the two reviewers exceeded 5, a third reviewer conducted a further evaluation and the final score of the article was defined as the average score of the third reviewer and a closer score given by either reviewer.
Descriptive statistics were reported for all the variables. Independent sample t-tests were used to compare the difference in CHEERS checklist scores by country of residence of the primary author, geographic location, and type of medical function. Analysis of variance was conducted to compare the difference in the CHEERS checklist scores by type of data collection and funding source. Pearson's correlation coefficient was performed to measure the relationship between the CHEERS checklist score, and the number of authors and the year of publication. Statistical analyses were conducted using SPSS version 21 and alpha level was set at 0.05.
RESULTS
The literature search identified 167 abstracts. After reviewing the abstracts of these articles, 150 articles were excluded because of: being multiple-country comparisons (n = 30), having no cost analyses (n = 21), being only a cost-of-illness study (n = 27), being a review article (n = 11), or being a veterinary study (i.e., involved animals with diseases) (n = 2), or did not compare pharmaceutical products (n = 59) (Figure 1). A total of seventeen studies were included for further evaluation (Table 1).
Flowchart of articles selection process.
Articles Evaluated in the Study
The earliest article was published in 1995, and the latest one was published in 2013. The seventeen articles were all published in different journals based in the United Kingdom (n = 7), the United States (n = 3), the Netherlands (n = 1), Italy (n = 1), Egypt (n = 1), Iran (n = 3), and Kingdom of Saudi Arabia (KSA) (n = 1). Most of the articles (n = 13; 76 percent) were published in medical journals, and the other articles were published in health/medical economic (n = 3; 18 percent) and pharmaceutical (n = 1; 6 percent) journals. Articles had an average of five authors (5 ± 2.7). Most of the primary authors (i.e., first authors) were from home countries (n = 13; 76 percent) and had medical or clinical training (n = 14; 82 percent). Nationwide studies accounted for 24 percent (n = 4) of the studies, and the rest were subnational studies (Table 2). Eight studies did not mention the perspective used (47 percent) and the most common perspective was the healthcare system (n = 4; 24 percent) followed by societal (n = 3; 18 percent) and payer (n = 2; 12 percent), respectively.
General Information for Included Studies
The reviewed studies conducted cost-effectiveness analysis (n = 5; 29 percent), cost-utility analysis (n = 5; 29 percent), cost-benefit analysis (n = 3; 18 percent), and cost-minimization analysis (n = 2; 12 percent). One study (6 percent) conducted both cost-effectiveness analysis and cost-utility analysis, and another study (6 percent) conducted cost-consequence analysis. The studies used modeling (n = 8; 47 percent), prospective (n = 6; 35 percent), and retrospective (n = 3; 18 percent) study designs for their analyses. All studies included direct medical costs, but only two (12 percent) included direct nonmedical costs, and two (12 percent) included indirect costs. In addition, six (35 percent) studies used primary data, ten (59 percent) studies used secondary data, and one (6 percent) study used both primary and secondary data. The studies were funded by universities (n = 6; 35 percent), author's own resources (n = 2; 12 percent), government (n = 2; 12 percent), private not-for-profit organizations (n = 2; 12 percent), and pharmaceutical industry (n = 1; 6 percent). However, the funding source was not listed in four studies (n = 4; 24 percent) (Table 3).
Economic Evaluation Information for Included Studies (n = 17)
The mean CHEERS checklist score for the seventeen pharmacoeconomic studies was 16 ± 4 of 24, ranging from 9 to 22. The interrater reliability of the reviewers was 75 percent. Although none of the CHEERS checklist score comparisons (Supplementary Table 1) were statistically significantly different, articles with first authors residing in foreign countries had a slightly higher mean CHEERS checklist score than did those residing in WHO EMC (17 versus 16); and national studies had a higher CHEERS checklist score than did subnational studies (19 versus 15).
DISCUSSION
This study summarized the reporting characteristics of pharmacoeconomic studies in WHO EMC that were available in English. Similar to other studies from Bangladesh (n = 12) (Reference Hoque, Khan and Hossain13), South America (n = 24) (Reference Machado, Iskedjian and Einarson15), Russia (n = 16) (Reference Gavaza, Shepherd, Shcherbakova and Khoza17), India (n = 29) (Reference Desai, Chandwani and Rascati4), Zimbabwe (n = 26) (Reference Gavaza, Rascati, Lawson, Lawson and Mann11), and China (n = 20) (Reference Jiang, Ma, Desai, Yang and Rascati18), this study had a low number of published full pharmacoeconomic studies (n = 17). This number is lower than the number of reviews from other countries such as India (n = 29) (Reference Desai, Chandwani and Rascati4), Zimbabwe (n = 26) (Reference Gavaza, Rascati, Lawson, Lawson and Mann11), Korea (n = 45) (Reference Lee, Brouwer, Lee and Koo10), Thailand (n = 41) (Reference Teerawattananon, Russel and Mugford16), China (n = 20) (Reference Jiang, Ma, Desai, Yang and Rascati18), and South America (n = 24) (Reference Machado, Iskedjian and Einarson15). The number of studies (n = 17) from twenty-two countries equals less than one study per country. This is far lower than the number of studies reported by others (at least 12 per country) (Reference Desai, Chandwani and Rascati4;Reference Lee, Brouwer, Lee and Koo10;Reference Gavaza, Rascati, Lawson, Lawson and Mann11;Reference Hoque, Khan and Hossain13;Reference Machado, Iskedjian and Einarson15–Reference Jiang, Ma, Desai, Yang and Rascati18).
The earliest study was published in 1995. There was an increase in the number of articles from 2008 to 2013. This is possibly due to the late arrival of pharmacoeconomics as a new discipline to the WHO EMC. In addition, the limited number of pharmacoeconomic studies in WHO EMC is likely due to the unavailability of retrospective data or difficulty in conducting studies in which primary data are collected.
To our knowledge, pharmacoeconomics is not one of the mandatory courses in many pharmacy schools and may be neglected in other schools in WHO EMC. A recent study found that of the pharmacy schools surveyed, half of them did not offer a pharmacoeconomics course (Reference Alefan, AlImam, Mukattash, Mhaidat, Alabbadi and Rascati20). Thus, the inadequacy of pharmacoeconomic education in WHO EMC could be another reason for the low number of publications. Lack of attention to health economic evaluations by healthcare decision makers could be another contributing factor. The overall extent of pharmacoeconomic studies in WHO EMC, however, is still limited compared with that in other countries.
At this time, there are no mandatory requirements for pharmacoeconomic assessments in most of the WHO EMC. Additionally, most of the countries in this region have no pharmacoeconomics guidelines like in many other countries (Reference Walker and Fox-Rushby7). Pharmacoeconomic evaluations can have a substantial impact on several aspects, such as drug pricing, promotion of health insurance reimbursement, rational use of drugs, and selection of essential drugs.
This study showed that less than one-third of the articles were published in domestic journals (n = 5; 29 percent). This rate is lower than that for other reviews where there were 33 percent (Reference Teerawattananon, Russel and Mugford16) and 31 percent (Reference Desai, Chandwani and Rascati4), but higher than that for others (10 percent) (Reference Jiang, Ma, Desai, Yang and Rascati18) and (13.6 percent) (Reference Gavaza, Rascati, Oladapo and Khoza12). This suggests that WHO EMC are not in a good position in comparison to other countries in the availability of domestic English language journals in which to publish pharmacoeconomic studies. Moreover, the remaining studies (n = 12; 71 percent) which are not published in domestic journals may not be reaching the intended audience in the region.
Although the majority (n = 13; 76 percent) of the first authors resided in WHO EMC, only approximately one-third (n = 5; 29 percent) published their studies in local journals. This might question the level of confidence that authors have in their countries' journals. Similar to other studies, cost-effectiveness (n = 5; 29 percent) (Reference Gavaza, Rascati, Oladapo and Khoza12;Reference Gavaza, Rascati, Oladapo and Khoza14;Reference Machado, Iskedjian and Einarson15;Reference Jiang, Ma, Desai, Yang and Rascati18;Reference Garattini, De Compadri, Clemente and Cornago21) and cost-utility (n = 5; 29 percent) (Reference Jiang, Ma, Desai, Yang and Rascati18) analyses were the most frequently used methods of economic evaluation. This indicates that the main questions/application in this area are answered by these two most common used analyses.
When compared with other reviews where 33 percent to 68 percent listed the funding source (Reference Desai, Chandwani and Rascati4;Reference Gavaza, Rascati, Lawson, Lawson and Mann11–Reference Teerawattananon, Russel and Mugford16;Reference Jiang, Ma, Desai, Yang and Rascati18;Reference Garattini, De Compadri, Clemente and Cornago21), the majority of studies in this review (n = 13; 76 percent) listed the funding source. However, in contrast to other reviews (Reference Desai, Chandwani and Rascati4;Reference Gavaza, Rascati, Oladapo and Khoza12–Reference Machado, Iskedjian and Einarson15;Reference Jiang, Ma, Desai, Yang and Rascati18), the most common source of funding in WHO EMC was the university (n = 6; 35 percent). This may indicate the low awareness and appreciation of the pharmacoeconomics discipline by governments in WHO EMC and that academics in universities are mainly the ones who are conducting pharmacoeconomic studies. This may be due to the absence of pharmacoeconomic guidelines and an official requirement for pharmacoeconomics studies in the region.
In general, the studies included in this review had an acceptable level of completeness (CHEERS checklist score = 16 of 24: fair). The general and economic characteristics did not significantly relate to the completeness level of studies, and this is maybe due to the relatively small sample size. The CHEERS checklist for healthcare-related economic publications was used based on the recommendation from previous study (Reference Jiang, Ma, Desai, Yang and Rascati18).
The study has several limitations. First, despite our best efforts, some published studies may have been missed or excluded. For example, the NHS EED was not searched. However, this risk should have been minimized by using systematic bibliographic search techniques. Second, we searched only the published literature in peer-reviewed journals, so other types of publications were not included in the search results. Third, only three scoring points per item were used. Although the inter-rater reliability was found to be adequate, other evaluators may score the articles differently.
CONCLUSION
This is the first systematic review of health economic studies conducted in WHO EMC. On average, the studies were of a fair level of reporting completeness. Appropriate allocation of limited healthcare resources is important in low and middle income countries. Hence, health economic evaluation should be encouraged in WHO EMC. Improving education and establishing guidelines may help to increase the number of health economic studies in this part of the world.
SUPPLEMENTARY MATERIAL
Supplementary Table 1: https://doi.org/10.1017/S026646231700037X
CONFLICTS OF INTEREST
The authors have no conflict of interest to declare.
