Financial incentives for return of service in underserved areas: a systematic review

Background In many geographic regions, both in developing and in developed countries, the number of health workers is insufficient to achieve population health goals. Financial incentives for return of service are intended to alleviate health worker shortages: A (future) health worker enters into a contract to work for a number of years in an underserved area in exchange for a financial pay-off. Methods We carried out systematic literature searches of PubMed, the Excerpta Medica database, the Cumulative Index to Nursing and Allied Health Literature, and the National Health Services Economic Evaluation Database for studies evaluating outcomes of financial-incentive programs published up to February 2009. To identify articles for review, we combined three search themes (health workers or students, underserved areas, and financial incentives). In the initial search, we identified 10,495 unique articles, 10,302 of which were excluded based on their titles or abstracts. We conducted full-text reviews of the remaining 193 articles and of 26 additional articles identified in reference lists or by colleagues. Forty-three articles were included in the final review. We extracted from these articles information on the financial-incentive programs (name, location, period of operation, objectives, target groups, definition of underserved area, financial incentives and obligation) and information on the individual studies (authors, publication dates, types of study outcomes, study design, sample criteria and sample size, data sources, outcome measures and study findings, conclusions, and methodological limitations). We reviewed program results (descriptions of recruitment, retention, and participant satisfaction), program effects (effectiveness in influencing health workers to provide care, to remain, and to be satisfied with work and personal life in underserved areas), and program impacts (effectiveness in influencing health systems and health outcomes). Results Of the 43 reviewed studies 34 investigated financial-incentive programs in the US. The remaining studies evaluated programs in Japan (five studies), Canada (two), New Zealand (one) and South Africa (one). The programs started between 1930 and 1998. We identified five different types of programs (service-requiring scholarships, educational loans with service requirements, service-option educational loans, loan repayment programs, and direct financial incentives). Financial incentives to serve for one year in an underserved area ranged from year-2000 United States dollars 1,358 to 28,470. All reviewed studies were observational. The random-effects estimate of the pooled proportion of all eligible program participants who had either fulfilled their obligation or were fulfilling it at the time of the study was 71% (95% confidence interval 60–80%). Seven studies compared retention in the same (underserved) area between program participants and non-participants. Six studies found that participants were less likely than non-participants to remain in the same area (five studies reported the difference to be statistically significant, while one study did not report a significance level); one study did not find a significant difference in retention in the same area. Thirteen studies compared provision of care or retention in any underserved area between participants and non-participants. Eleven studies found that participants were more likely to (continue to) practice in any underserved area (nine studies reported the difference to be statistically significant, while two studies did not provide the results of a significance test); two studies found that program participants were significantly less likely than non-participants to remain in any underserved area. Seven studies investigated the satisfaction of participants with their work and personal lives in underserved areas. Conclusion Financial-incentive programs for return of service are one of the few health policy interventions intended to improve the distribution of human resources for health on which substantial evidence exists. However, the majority of studies are from the US, and only one study reports findings from a developing country, limiting generalizability. The existing studies show that financial-incentive programs have placed substantial numbers of health workers in underserved areas and that program participants are more likely than non-participants to work in underserved areas in the long run, even though they are less likely to remain at the site of original placement. As none of the existing studies can fully rule out that the observed differences between participants and non-participants are due to selection effects, the evidence to date does not allow the inference that the programs have caused increases in the supply of health workers to underserved areas.


Time series
All students who ever participated in the program between 1953 and 1977 and who completed their medical training in 1975 or before (N = 149)

Records of the Western Interstate Commission for Higher Education
Proportion of participants who had started serving or completed their practice obligation in 1975 or before: Of 149 participants, 67 (45%) served the obligation in a metropolitan area within Arizona, 21 (14%) served the obligation in a non-metropolitan area in Arizona, and 55 (37%) repaid the financial incentive, while the remainder defaulted.

Time series of medical student density in Arizona:
The per-capita number of medical students did not increase from 1953 to 1967 (consistently 20% below national average), but increased steeply from 1968 onwards.
About two fifths of participants did not fulfill their obligation to practice in an underserved area.
85% of participants who completed their obligation remained in Arizona.
The program did not succeed in increasing the medical student population density in Arizona. The steep increase in per-capita medical students in 1968 is attributed to the opening of the first medical school in Arizona in that year.  In communities in northern Ontario with a population of less than 15,000 the relative number of physicians declined slightly from 1966 to 1969 (i.e., before the program was introduced), while it increased steeply from 1970 (after introduction of the program) to 1972 (from 1.0 to almost 1.3).
About half of participants did not fulfill their obligation to practice in an underserved area.
About three quarters of participants who completed their obligation remained at the original placement location.
The time series suggests that the program was effective in increasing the number of physicians practicing in small communities in northern Ontario.   Proportion of participants who intended to leave their NHSC practice location who provided the following reasons for leaving: "The most commonly cited reasons for leaving (each respondent could give two primary reasons) were: 1) the site was geographically isolated or was unpleasant in some other way (61 percent); 2) salary at the NHSC site was insufficient (31.5 percent); 3) on-call and clinical responsibilities associated with the NHSC position were excessive (28.5%); and 4) children's needs or spouse's career or other needs were unmet (26.2%)" [14].
Only about one third of participants who were currently fulfilling their practice obligation intended to remain in their placement practice after completing their obligation.
The major reasons for intending to leave the placement site were dissatisfaction with the community, the salary, and the workload, as well as unmet needs of family members. allopath", "physician specialty ("family medicine vs. other)", "physician initial underserved-area retention plans (10 years or longer vs. less than 10 years)", "practice percentage of minority patients", "county population", county status (metropolitan vs. nonmetropolitan)", "county per capita income", and county primary care physician-to-population ratio" Mean satisfaction score: NHSC physicians: Across 18 dimensions of satisfaction, 9 mean satisfaction scores were between "dissatisfied" and "neutral", 8 were between "neutral" and "satisfied", and one ("Caring for needy patients") was between "satisfied" and "very satisfied". Across 15 dimensions of satisfaction, NHSC physicians reported significantly lower satisfaction than non-NHSC physicians for "Referral Access to Consultants", "Freedom from Bureaucratic Interference", "Clinical Autonomy", "Opportunities to Achieve Professional Goals", "Earnings From Practice", "Quality Physician-Patient Relationships", "Life in a Small Community", "Climate or Geography", "Access to Cultural Activities", "Having Amenities of City Living" (all p ≤ 0.006), and reported higher satisfaction only for "Caring for Needy Patients" (p = 0.003). Mean satisfaction score: minority vs. non-minority NHSC physicians: Across 17 dimensions of satisfaction, all mean satisfaction scores for both minority and nonminority physicians were between "dissatisfied" and "satisfied", with the exception of the score for "Caring for needy patients" which was between "satisfied" and "very satisfied" for both groups.
Minority physicians had significantly lower mean scores than their non-minority peers for "Clinical autonomy", "Opportunities to achieve professional goals", "Earnings from practice", "Opportunities for outdoor sports", "Life in a rural community", "Climate or geography", and "Ability to find compatible friends" (all p ≤ 0.04). The positive association of participation with practice in rural areas and with the proportion of Medicaid and uninsured patients remained significant "while controlling for selected characteristics of physicians".
In comparison to nonparticipants, participants in financialincentive programs were about five times more likely to practice in rural areas and 85% more likely to care for underserved populations.
Observational study Selection bias due to selective participation in financialincentive programs not controlled for