The dismal state of tuberculosis (TB) care in the private sector in India has been exposed by a November 2015 study (The Lancet) that employed a novel strategy of using standardised TB patients — healthy individuals trained to pose as TB patients and interact with health-care providers. Standardised patient method has been widely used in India for several other conditions such as asthma, angina, diarrhoea, and pneumonia.

“Relative to other methods [questionnaire surveys, recall-based patient surveys, analysis of prescription and patients’ medical records], data from standardised patients yield an assessment of provider practice free from observation bias, less vulnerable to recall bias, and more complete than medical records. Furthermore, standardised patients permit estimates of case detection rates since illnesses are fixed by design. Finally, because cases are standardized, the methodology allows for quality comparisons across different care providers (public and private practitioners),” Dr. Madhukar Pai, the senior author of the paper from McGill University, Montreal, Canada, says in an email to me.

Seventeen healthy individuals who were trained to pose as TB patients made 250 interactions with 100 doctors who had consented to the study. The 17 individuals interacted with doctors as either patient 1 or patient 2 (presumed TB — suspected but not already diagnosed to have TB), patient 3 (confirmed TB) or patient 4 (suspected MDR-TB). The doctors practising in low- and middle-income settings in Delhi had a MBBS degree (29 numbers), a degree in alternative medicine (40 doctors) or were informal health-care providers with no qualification. All the interactions took place between April 1 and April 23, 2014.

The standardised patients were correctly managed for their respective TB conditions only in 21 per cent (52 of 250) interactions.  The ones who presented with confirmed TB were the most correctly managed ones — 14 per cent for treatment and 26 per cent for referral to DOTS centre or qualified providers. In contrast, correct case management was seen in only 12-13 per cent in the case of standardised patients with presumed TB.

In the case of suspected MDR-TB, referral to a DOTS centre or a qualified provider was only 20 per cent, indicating a “low use of drug-susceptibility testing”.  Contrary to the general notion, MBBS doctors were less likely to correctly treat those with suspected MDR-TB as they started treatment without seeking drug-susceptibility testing.

MBBS doctors were more likely to correctly manage patients and more likely to ask for sputum tests and chest X-rays than others.Like earlier studies that revealed private practitioners’ tendency to initially treat TB patients with antibiotics, the current study found this practice to be widely prevalent. While fluoroquinolones were given to all the four standardised patients in 10-16 per cent of cases, broad-spectrum antibiotics were widely given to those who presented with presumed TB.  Even in the case of standardised patient 2 who had presumed TB but had already finished a round of antibiotic treatment, the doctors were less inclined to ask for sputum test or chest X-ray; steroids were given to them in at least 13 per cent of all interactions.

Microbiological confirmation is the gold standard for TB diagnosis. In spite of having such a microbiological confirmation, standardised patients with confirmed TB and suspected MDR-TB were asked to take a chest X-ray by a large proportion of the doctors. Chest X-ray is a screening tool and has less relevance when microbiological confirmation is already available.

The study revealed that MBBS doctors were more likely to correctly manage “patients” and more likely to ask for sputum tests and chest X-rays than others. Yet, they were “equally likely” to prescribe antibiotics and “more likely” to prescribe fluoroquinolones than others.

Deviations and delay

Incidentally, the study found a major gap in provider knowledge and practice. While doctors showed low adherence to established standards of TB care in clinical practice, they had “markedly high levels of knowledge”.

The private sector is the first point of contact for nearly 80 per cent of TB patients in India, and it offers TB care to nearly 50 per cent. Yet, the deviation from the established standards of TB care in clinical practice, as seen in this study, is the reason why TB patients most often encounter a delay of nearly two months before a correct diagnosis is made and are subjected to high levels of treatment failure and drug resistance.

The doctors were able to detect the standardised patients only in 11 of 232 (five per cent) instances though they were informed before hand that they might come across “patients” who were not real.

“Our study was a pilot because we ourselves did not know whether they [standardised TB patients] would go undetected. We were told that standardised patient method will not work for TB because patients need to have fever and appear sick. But that is not true — unlike malaria or pneumonia, TB fever comes and goes, and is often low grade (night sweats). And TB patients do not look very sick (unless they have advanced TB or MDR-TB). So, we were happy that the detection rate was [only] about 5 per cent,” says Dr. Pai. “Now that we know it works, we are doing much larger studies.”

---

Related story and link:

Trials with antibiotics lead to TB diagnosis delay

 

Published in The Hindu on July 31, 2016