100060 from EVIDENCE to POLICY Learning what works for better programs and policies September 2015 ARGENTINA: Can Short Term Incentives Change Long Term Behavior? It’s hard to change how organizations do things, whether The World Bank partners with governments to provide it’s how they build a product or how they provide service innovative solutions for ensuring that pregnant women to a patient. Even when there are guidelines showing best and their young children get the care they need to grow practices, there’s always up healthy. Knowing what works—and what doesn’t— is a cost to instituting critical to succeeding. In Argentina, the World Bank sup- new routines. There’s ported the government’s Plan Nacer health program that also a cost to not mak- gives provincial authorities financial incentives for enrolling ing changes, especially pregnant women and children and for meeting specific pri- when it comes to health mary health goals. As part of this, researchers tested whether care. Prenatal care start- providing a temporary increase in financial incentives to ed at the right time, for clinics would encourage providers to initiate care for preg- HEALTH example, can improve nant women in the first trimester—and whether this would the odds of a healthy continue even after the increase stopped. The evaluation— birth. Started late, or not provided at all, the mother and the first of its kind to examine the effects of short-term in- her baby are at risk. Pay-for-performance schemes have suc- centives on long-term performance—found that the boost cessfully improved the quality and timing of healthcare, but worked and the better care continued even after the incen- maintaining such bonuses over the long term is costly and tives ended. The results provide valuable insights into too expensive for many health budgets. Are there better the possibility of using temporary incentives to change ways to promote new practices? behavior over the long term. Context After an economic crisis in Argentina in 2001 plunged mil- this way led to an improvement in health care, as measured lions of people there into poverty and worsened health out- by a drop in low birth weight babies and a decline in new- comes, especially for women and children, the government born deaths. (See Evidence to Policy Note ARGENTINA: crafted a new, national plan for a provincial health insurance Can Performance Payments Improve Newborn Health?) program for the poor. The program, called Plan Nacer, was Despite improvements in health care and outcomes, specifically designed for pregnant women without health only a third of pregnant women covered by Plan Nacer were coverage and children up to the age of six (the program has coming in for their first prenatal visit in their first trimester, since been expanded to include other groups). Under the an important time for health workers to identify problems program, which was launched in 2005, the national govern- and advise women on proper nutrition. The National Min- ment transfers money to the provinces to use for health ser- istry of Health, jointly with the provincial authorities and vices, basing part of the payment on how well the province the World Bank, decided to test whether temporarily offer- has done at meeting certain health indicators. An evaluation ing clinics higher payments for starting prenatal care in the of the program found that setting up the financial incentives first trimester would help. The eight-month intervention was implemented in been operating in Misiones for five years. Under the pilot, primary health care clinics in Misiones province, one of Plan Nacer clinics assigned to be part of the evaluation the poorest in the country and a place where pregnant agreed to a new fee structure based on getting women women and young children are most at risk of dying. into clinics for prenatal treatment before the start of their When the pilot was launched in 2010, Plan Nacer had 13th week of pregnancy. Evaluation Of the 262 primary care clinics in Misiones province, 37 for a subset of pregnant women. To figure what week of preg- were included in the evaluation. Eighteen of these clinics nancy women were in when they made their first prenatal were randomly assigned to the treatment group and 19 were visit, researchers measured the difference between the first assigned to the control group. Clinics in the treatment group prenatal visit and the patient’s last menstrual date, which is followed a modified fee schedule for prenatal visits that paid used by health care workers to calculate a baby’s due date. them $120 ARS (about US $31 based on the exchange rate External auditors verified the clinical records, and researchers at the time) if the first prenatal visit occurred before the 13th cross checked hospital records to verify gestational age at birth week of pregnancy. Clinics received $40 ARS (about US and rule out the possibility that clinics misreported the stage $10.30) for each subsequent visit. The control group clinics of pregnancy to secure the additional incentives. continued on the regular fee schedule, under which clinics Data was separated into four groups: before the pilot was were paid $40 ARS for any prenatal visit—regardless of when implemented from January 2009 to April 2010; when the HEALTH in the pregnancy the visit took place. Because of compliance pilot was implemented from May 2010 to December 2010; issues, however, only 14 clinics actually received treatment, a follow up period between January 2011 and March 2012, and one of the control clinics was offered treatment acciden- after the pilot had ended; and then a final follow-up pe- tally. Another clinic closed before the intervention began. riod from April 2012 through December 2012. Because the Public health clinics in the province have an automated province instituted a new information system for data start- health records information system, which include dates of ing in April 2012, it wasn’t easy to compare data available visits, services delivered, and birth weights. Using patients’ in the last evaluation round to data available previously. As national identity number, researchers were able to merge a result, the impact of the program after incentives were re- health data with the program beneficiary status. moved is harder to measure at the two-year follow up mark. Researchers focused on services for pregnant women who Researchers also conducted in-depth interviews with were Plan Nacer beneficiaries at the time of their first prenatal health care workers to better understand what methods they visit, but information on birth outcomes was only available used to initiate early prenatal care. Findings Giving clinics higher payments for prenatal visits than women who went to clinics that were paid the same that occurred before the end of the first trimester fee regardless of when the visit took place. Overall, 42 improved the rate of early prenatal care. percent of pregnant women treated by health clinics that qualified for the higher payments had their first prenatal In health clinics that followed the new fee structure, preg- visit before their 13th week of pregnancy, compared with nant women had their first prenatal visit 1.5 weeks earlier 31 percent of women in clinics that followed the tradi- This policy note is based on World Bank Policy Research Working Paper 7348, “Long Run Effects of Temporary Incentives on Medical Care Productivity,” Pablo Celhay, Paul Gertler, Paula Giovagnoli, Christel Vermeersch; June 2015. The impact evaluation was supported by the Health Results Innovation Trust Fund (HRITF) and the Strategic Impact Evaluation Fund (SIEF). The working paper can be found at: http://documents.worldbank.org/curated/en/2015/06/24736285/long-run- effects-temporary-incentives-medical-care-productivity tional fee payment schedule. The rate of initiating prenatal The new focus on getting pregnant women in for care in the first three months rose by 34 percent in clinics prenatal care didn’t lead clinics to cut back on that followed the new structure. other services they provided. Even after the incentives ended, clinics that There was no evidence of any negative spillover onto oth- had received the higher fees continued to have er services, such as the overall number of health visits by improved rates of early prenatal visits among women and children. For example, before the pilot program their pregnant patients. started, more than 80 percent of pregnant women received a tetanus vaccine. The figure stayed constant during and after The higher levels of early care persisted for at least 15 months the pilot. and likely continued for more than 24 months after clinics stopped receiving the higher fees for prenatal care that happened in the first trimester. Long after the higher payments stopped, women in the treatment group were still starting prenatal care 1.6 weeks before women in the control group. Overall, there was an eight percentage point difference in the rate of first tri- mester care between women seen by clinics that had received extra fees and those seen by clinics that were paid as usual. The higher payments resulted in changes in medical care routines, as clinics devised creative ways to identify women in the first trimester of pregnancy and to encourage them to come in early for prenatal care. The higher fees motivated clinics to create new strategies for making sure pregnant women had their first prenatal care visit before the 13th week of pregnancy. In some clinics, for Early prenatal care is important for the health of example, individual team members received bonus payments mother and child—but the pilot didn’t lead to any based on the number of pregnant women they brought in ev- drop in the rate of premature birth or increase ery month. In other clinics, staff would check whether female birthweight. patients were picking up their birth control pill prescriptions and then prioritize home visits to those who hadn’t. Researchers believe there may be a number of reasons for the Clinics also tried to reach out directly to women who lack of any measured impact on birthweight and premature already had children, since they were more likely to de- births. First, the sample was relatively small to detect any lay the first prenatal visit, as compared with women who statistically significant change. Second, it may be that best were pregnant for the first time. Since these women often practices matter when it comes to prenatal care, but mainly were eligible for free milk for their older children, health for women who are high-risk because of substance abuse, workers would talk to them during the milk distribution, smoking, pre-existing conditions or because they don’t see a ask when they had last menstruated and give the women doctor until close to the end of their pregnancy. It’s possible instant-read pregnancy tests to those whose menstruation that high-risk women weren’t the ones who changed their was late. In some clinics, workers realized that adolescents behavior and came in for prenatal care in the pilot—but were particularly unlikely to receive early care, since their creating a more targeted program is difficult. pregnancy was often a secret from their parents, and so health workers changed the timing of home visits so that Still, the pilot showed that it’s possible to get they’d be more likely to find the teenager home alone. health clinics to change how they do things— These new and expanded outreach methods were sus- and that the new behavior sticks even after tained, even after the incentives stopped. incentives are removed. The incentives convinced health care workers to overcome costs associated with lowered productivity while the health their resistance to change, which researchers believe to be care workers learned the new routines, making it worth- one of the biggest barriers to compliance with clinical care while for clinics to institute changes. guidelines. The increased fees probably helped cover the Conclusion As the results of this impact evaluation show, tempo- Nevertheless, the results also highlight the challenge of rary incentives are an effective way to motivate health improving birth outcomes for high-risk populations. As care workers to change their routines so that ultimate- researchers continue to search for innovative ways to ly, they’re providing better care for the people they’re help the world’s poor, the lessons from this intervention tasked to serve. The results will be particularly useful to underscore the importance of providing solutions that policy makers looking to make long-term changes more specifically target those who need it most. cheaply than traditional pay-for-performance programs. HEALTH The Strategic Impact Evaluation Fund, part of the World Bank Group, supports and disseminates research evaluating the impact of development projects to help alleviate poverty. The goal is to collect and build empirical evidence that can help governments and development organizations design and implement the most appropriate and effective policies for better educational, health and job opportunities for people in developing countries. For more information about who we are and what we do, go to: http://www.worldbank.org/sief. The Evidence to Policy note series is produced by SIEF with generous support from the British government’s Department for International Development. THE WORLD BANK, STRATEGIC IMPACT EVALUATION FUND 1818 H STREET, NW WASHINGTON, DC 20433 Produced by the Strategic Impact Evaluation Fund Series Editor: Aliza Marcus Writer: Daphna Berman The original had problem with text extraction. pdftotext Unable to extract text.