Connecting the dots from detection to cure
If governments don’t have a good picture of the health care needs of their citizens, how can they begin to address them effectively and efficiently? Governments around the world are all too familiar with this paradox, which can be particularly challenging if administrative data on health care delivery and outcomes is of poor quality or out-of-date.
These issues have been magnified in the current COVID-19 pandemic, and will continue to be as practices, like contact tracing, become a larger part of health system responses. Unfortunately, they are not new and have played a role in the spread of many other diseases.
This includes Tuberculosis (TB), which continues to be a persistent disease: 10 million people contracted TB in 2018 and 1.5 million died.
To control its spread, health care providers must detect infection early, ensure patients adhere to a strict treatment regimen, and monitor and report patient progress to generate data that feeds into the larger health system. Though seemingly straightforward, monitoring patients to ensure they complete treatment and health care providers to ensure they deliver quality care is a daunting task.
Many TB programs struggle to connect the dots from detection to cure. India is no exception, accounting for 27 percent of the global TB disease burden.
Medical records are often paper-based. Tracking patients across multiple registers to see if they’re adhering to treatment is a time-consuming and error-prone task for already capacity constrained health providers.
Absent proper monitoring by management, compensation schemes based on number of new cases or patient outcomes may wrongly incentivize providers to falsify data. Close oversight of provider performance and the accuracy of their reporting, however, is extremely difficult because service delivery is highly decentralized.
So, how do countries, like India, design large scale monitoring systems to better piece together the picture of health care delivery and outcomes in real time?
To address this question, J-PAL affiliate Vincent Pons, along with Thomas Bossuroy and Clara Delavallade of the World Bank, tested the impact of switching from paper-based patient tracking to electronic biometric tracking on patients’ health outcomes, provider performance, and the accuracy of patient data.
Researchers partnered with Operation ASHA (OpASHA), India’s largest nonprofit TB care provider. They randomly assigned 65 areas served by OpASHA centers or mobile workers to either receive biometric technology, or serve as the comparison group and not receive the intervention.
At TB centers with the technology, patients and providers scanned their fingerprints each time they arrived at the TB center, and the data was sent daily to OpASHA’s server. The technology aimed to: 1) ensure patients themselves received medication from DOTS centers; 2) automatically alert health workers when patients failed to show up for treatment; and 3) enable OpASHA management to review health worker effort and performance in real-time.
The research team found that biometric tracking increased patient adherence to TB treatment. Patients were less likely to interrupt treatment and were also more likely to consume their medication at TB centers, as opposed to sending relatives to collect it on their behalf.
Health worker effort was a major factor behind these results. Provider attendance and time spent at centers increased, as did the frequency of home visits to follow-up with defaulting patients that had been flagged by the technology. Health workers had a better picture of patients’ progress through the health care system, and OpASHA management potentially had more stringent oversight.
Tracking also improved the accuracy of data relative to the comparison group. At OpASHA, performance is measured by seven indicators, including the number of new TB cases identified and the number of patients completing their treatment. Most health workers receive cash bonuses for each new case detected, and, without proper monitoring, can be tempted to falsify data to meet these targets.
Because it reduced the opportunity for health workers to alter health records, biometric devices decreased overreporting of new cases and underreporting of treatment interruptions. Despite reducing cash bonuses paid to health workers because of more accurate recordkeeping, their job satisfaction remained the same. The devices cut provider workloads because they no longer had to manually track patient progress across multiple paper records.
What the picture shows
Many evaluations have identified effective strategies targeted at patients to improve health outcomes, like behavioral nudges to improve treatment adherence via SMS or information on pill packaging. However, interventions that improve provider effort, such as through better monitoring of their performance as this evaluation suggests, can be a viable strategy to boost medication adherence.
Other evaluations from India show that monitoring attendance through technology improved teacher and health care provider performance. Yet, monitoring systems can be undermined over time if there is not broader organizational support. In this evaluation, monitoring devices were not met with strong resistance, perhaps because they reduced provider workloads by facilitating patient tracking.
As this evaluation suggests, biometric tracking has the potential to improve accountability and overall quality of health systems by providing a more accurate and more timely picture of whether services are being delivered and whether they are translating into outcomes. India’s existing infrastructure—Aadhaar, which has provided almost nationwide coverage of biometric identification cards to residents, and Nikshay, its online database of all TB patients—presents opportunities that are ripe for scaling.
This could have far reaching implications for strengthening government capacity, beyond the important case of TB control. Technology, as this evaluation has shown, can help governments better connect the dots between service delivery and outcomes by improving the reliability of admin data, providing providers and management an accurate picture of service provision in real time, and reducing the burden on frontline workers.