Fighting TB in India’s Cities

1 November 2018
Fighting TB in India’s Cities
READ THE FULL BLOG BY JISHNU DAS, BENJAMIN DANIELS, ADA KWAN AND MADHUKAR PAI

After an autopsy lasting 200 years scientists concluded that the mummy Irtyersenu, first revealed to London’s Royal Society in 1825, died of tuberculosis (TB) around 600 BCE. Today, TB still causes much suffering and lost productivity around the world, despite the perception of ‘no longer being an issue’ in higher income countries. In 2017, TB affected 10 million people, and killed more people than HIV/AIDS – a total of 1.3 million fatalities worldwide.

India may well be the `epicentre’ of the disease: The country accounted for a quarter of the global TB cases and TB deaths in 2017. Nearly a third of the world’s 3.6 million undiagnosed or unreported cases are thought to live there, and in Mumbai, the first strain of the TB bacteria resistant to all known treatments was reported in 2012.

The good news is that we know how to control the epidemic and bring it close to eradication. Well-executed public health programs that screen potential patients, bring them to treatment quickly, and ensure treatment completion can halt the spread of this ancient scourge. In China, the World Health Organisation estimates that vigorous efforts to bring treatment to those already diagnosed halved the TB prevalence rate of TB from 215 per 100,000 people in 1990 to 108 in 2010.

But India is different. It was long conjectured that, like in China, the majority of TB patients were receiving care in the public sector. But then, in 2016, Nimalan Arinaminpathy and colleagues used an ingenious method to come up with realistic estimate. Based on the sales of anti-TB drugs in the private market, they showed that, in fact, it was the private sector that was treating two-thirds of India’s TB patients.

There were 17.793 million patient-months (…) of anti-tuberculosis treatment in the private sector in 2014, twice as many as the public sector. If 40–60% of private-sector tuberculosis diagnoses are correct, and if private-sector tuberculosis treatment lasts on average 2–6 months, this implies that 1.19–5.34 million tuberculosis cases were treated in the private sector in 2014 alone. The midpoint of these ranges yields an estimate of 2.2 million cases, two to three times higher than currently assumed.’

Given that most prior anti-TB programs had been conceived of and delivered in the context of a top-down public healthcare system, how would India manage in this completely differently context?

A new approach to combating TB – with the private sector

In 2014, the Indian RNTCP, Mumbai Corporation and Bihar State government, in partnership with NGOs PATH and World Health Partners, with support from the Bill and Melinda Gates Foundation India, started to work on this difficult challenge. As part of a new program, World Health Partners in Patna and PATH in Mumbai, serving the role of intermediary agencies, implemented Private Provider Interface Agencies that sought to expand the capacity and improve the quality of TB diagnosis, treatment, and care in the private sector. This was not easy.

While prior studies hinted at low knowledge about TB standards in the private sector and long diagnostic delays, there was lack of data on how exactly private practitioners managed TB. What systematic mistakes were providers making? Were there pockets of excellence in these cities? How could outstanding doctors be identified? How do informal and AYUSH providers manage TB?

To help implement this program, our team, we put in place the world’s largest surveillance of TB care quality in those two cities. Instead of focusing on administrative data that has severe limitations in these contexts, we decided to use a gold-standard method of quality measurement for primary care: standardised patients, or SPs. Our methodology was first validated in a pilot study in Delhi, and has been subsequently used for our larger study in India, as well as similar studies in Kenya, China and South Africa.

{A brief interlude. In an SP interaction, a trained professional pretends to be a patient and visits a doctor. The SP presents with a pre-determined set of symptoms, a scripted personal history, and pre-planned responses to common questions. The same SP may present that scenario to hundreds of different doctors during one study. Since those presentations are all identical and researchers know what the SP was presenting with, the quality of care the SP received can be benchmarked to standards of care and accurately compared across providers and across time. Furthermore, since the provider does not know their actions are being recorded, the SP approximates the quality of care delivered to real patients.}

Together with the Institute for Socioeconomic Research on Development and Democracy, we recruited and trained 24 people from the local community, who then took on the tough task of presenting as SPs to multiple providers. After a month-long training process, our brave SPs went on to complete 2,652 interactions at 1,203 health facilities. Because the two organisations in these cities had painstakingly completed street-level listings of healthcare providers, we were able—for the first time—to create representative estimates of quality in 2014-15, just before the programs scaled up for implementation.

Quality in the private sector

Our recently published paper reports the results from our pre-program study in 2014-15, summarising the quality of care delivered by private health care providers in these cities at that time. There were three key features of the ‘market’ for TB care in these cities.

First, patients choosing a health care facility at random would have been treated in a manner consistent with national and international guidelines just 35% of the time. We believe that even this is an overestimate because we did not penalise providers when they also gave patients unnecessary (or even harmful) tests and medications. In fact, most patients also received unnecessary medications, including broad-spectrum antibiotics that may contribute to drug-resistance, or fluoroquinolone antibiotics and steroids which may mask symptoms while the TB infection worsens. The fraction of providers who correctly managed the case without giving unnecessary medicines is a frightening 9%.

But here is the thing. These low numbers are not because `all providers are low quality’ but instead reflect the tremendous quality variation in these cities. Part of this variation is because medical care in these cities is delivered by providers ranging from highly specialised chest specialists and MBBS doctors to those with degrees in Ayurveda, Yoga, Unani, Siddhi and Homeopathy (collectively called AYUSH), as well as providers without any formal qualifications at all. In Mumbai, 50% of providers in program areas were AYUSH; in Patna, 40% had no formal medical training at all. And MBBS providers consistently provided higher quality care, correctly managing 46% of cases compared to 23.5% for AYUSH and informal providers.

But part of this variation is also because, within both groups, there was always some providers who correctly managed every SP and some who got every case wrong. Providers were remarkably consistent — those who managed TB right once got it right again. Those who wrongly gave steroids and antibiotics to TB patients did so repeatedly.

The second feature of the market was therefore the recognition that there were excellent doctors in both cities, but qualifications were only a crude marker of quality.

Third, we are all accustomed to hearing about over-medication among healthcare providers. This is evident in our data, but there are also more nuanced patterns. The good news is that only 3% of our SPs received anti-TB medications without a microbiological test result, and that anti-TB medications were almost exclusively given by providers with the appropriate qualifications. Overuse of anti-TB medication had been a major concern for TB control programs, and we have now been able to show that neither pharmacists nor informal and AYUSH providers abuse anti-TB medications. The bad news is that all types of providers continue to use antibiotics indiscriminately, and more worryingly, some providers prescribe fluoroquinolones and steroids, both of which can mask the symptoms of TB and render diagnosis more difficult.

Scaling up anti-TB programs: A four-fold path to success?

Equipped with this granular description of the health market, the programs in these cities started to scale up in 2014. After 2015, both Mumbai and Patna started to see significant improvements in TB notification rates among private sector providers with greater use of microbiological tests and improved treatment completion rates. We have now begun to analyse new data from the quality of care surveillance conducted after the program was in place (not included in our recent publication). Our preliminary results suggest the program accomplished substantial improvements in how patients were being managed. Although these are early days, the signs are encouraging.

Now, the government, with support from The Global Fund, is expanding this model of private sector engagement to several cities through its Joint Effort for Elimination of Tuberculosis. And it is likely that they will face a comparable situation, with high quality dedicated doctors practicing amidst many indifferent and mediocre providers.

Based on our experience, we propose a strategy called IFMeT that may be key to successful private-public partnerships to fight TB. The four components that comprise IFMeT are Identification, Focusing, MEssaging and Testing.

Identification: Our data show that there are great providers in every city, but there are many providers with doubtful quality. Identifying `champion’ high quality providers early in the program is critical to the success of the program. The SP method is one tool in the arsenal, but care is required to balance anonymity and program success.

Focusing: Quality improvement efforts often work with a large set of providers through trainings and continuing medical education. An alternative to improving average quality is to connect patients with pre-identified top providers and focus investments and training on this small provider group, while leaving lower-volume or lower-quality providers untouched. In both Mumbai and Patna, 20% of providers were responsible for 80% of patient volume initially, suggesting high returns to focused effort. This `provider focusing’ approach massively decreases the scale of the program while retaining virtually all its benefits.

Messaging: The widespread use of unnecessary medicines may be tied to financial incentives. But financial linkages in the private sector are complex, closely guarded and difficult to uncover or build a program around. A better approach is to concentrate in one or two key behaviors. At this point, the indiscriminate use of anti-TB drugs is no longer of paramount concern. Fighting the use of broad-spectrum antibiotics may have to wait. But the use of fluoroquinolones and/or steroids can make TB diagnosis even harder and messaging only on the overuse of fluoroquinolones and steroids can have substantial impact.

Testing: We often assume that the main problem in the private sector is over-testing driven by profit motives. In fact, most providers are getting diagnoses wrong because they don’t test enough. Doctors need to do more X-rays, sputum tests and GeneXpert tests for patients presenting with symptoms consistent with TB, even when the odds of a positive result seem low. We found that when doctors were given better diagnostic information like test results, their decisions were more appropriate, and they gave fewer unnecessary medicines. Encouraging more testing, whether through price reductions for the patient and private laboratories, or communication with the provider is key to success when it comes to TB.

IFMeT, these strategies will take a very large problem without clear boundaries and bring it down to a series of actionable and manageable steps. Granville’s Egyptian mummy suffered many ignominies throughout her history, not the least of which was 200 years of being assigned a wrong cause of death until DNA testing revealed TB to be the culprit. With the resurgence of TB, the rise of drug resistant strains and the fact that this disease, more than any other, lays bare a society’s claims to fairness and equity, we cannot afford such mistakes in diagnosis any longer.

For additional resources: Quality of TB Care (QuTUB) Project https://www.qutubproject.org/

Jishnu Das is a Lead Economist at the World Bank, Washington DC and a visiting Senior Fellow at CPR.

Benjamin Daniels is a researcher at the World Bank, Washington DC.

Ada Kwan is a doctoral candidate at the University of California, Berkeley.

Madhukar Pai is Director of the McGill International TB Centre, Montreal.

The views shared belong to individual faculty and researchers and do not represent an institutional stance on the issue.