Diagnosis in ‘Digital India’

Divvy K. Upadhyay Dean F. Sittig Hardeep Singh

November 24, 2014

The HinduNO ROOM FOR ERROR: "The opportunity to diagnose Ebola correctly was missed in Thomas Duncan’s case in the U.S., but there are several lessons to be learnt." Picture shows wards at a hospital in New Delhi. File photo: Meeta Ahlawat

The government must recognise the role low-cost health IT innovations could play in improving diagnostic accuracy, including many that would be useful for rural India

The diagnosis of the first patient with Ebola in the U.S. was initially missed in an emergency room late night on September 25. Thomas Duncan, a Liberian national visiting Dallas, Texas, complained of flu-like symptoms and fever, but after lab work and CT scans, was given antibiotics and discharged with presumed sinusitis. The opportunity to diagnose Ebola correctly was missed. But there are several lessons to be learnt, many of which are relevant for ‘Digital India’.

Duncan informed a nurse about his travel history from West Africa. She documented this in the electronic health record (EHR). However, this did not set alarm bells ringing despite Ebola awareness preparation preceding this event. And then the hospital fumbled on its explanation, first blaming the nurse saying she did not tell the doctors about the travel history and then blaming a glitch in the EHR saying that because the nurse’s workflow (where the travel history was recorded) was not aligned with the doctor’s workflow, the doctor did not have access to the nurse’s notes. The next day it backtracked saying the EHR system worked just fine. At a U.S. Congressional hearing three weeks later, the hospital finally admitted misdiagnosis. This was from a hospital that was personally chosen by none less than former U.S. President George Bush for a cardiac stent procedure last year.

Missing critical information

Digitisation is aimed at improving health care delivery, quality and safety, and also facilitating the measurement of quality and safety metrics. While there was no computer glitch, the timeline released by the hospital suggested that the doctor spent more time with the computer than with the patient. Doctors in the U.S. now routinely use templates and pick-lists to help them navigate the complexities of medical documentation, billing and quality reporting requirements. But complex human-computer interactions in busy and chaotic clinical settings can also lead doctors to miss critical information. Humans become more reliant on technology and less likely to communicate face-to-face. In this case, the nurse ideally should have let the doctor know verbally about the travel history. Large amounts of data in the EHR can lead one to miss critical “signal” information versus the “noise.” We do know that basic clinical interactions with patients often suffer when attention is diverted to the computer so much.

This case is a perfect prototype for one of the biggest vulnerabilities of medicine — misdiagnosis. In the U.S. alone, 1 in 20 adults are estimated to be misdiagnosed annually in outpatient settings. The numbers are likely to be higher in India. Errors in diagnosis of several common diseases have been well described. Many of these diseases are common in India, including infections, cancers and cardiovascular conditions. Both problems in clinical judgment such as failing to gather or synthesise history, a physical exam or test data appropriately, as well as healthcare system flaws such as a lack of record or data availability, problems with communication or coordinating care, and insufficient access to specialists, can all contribute to misdiagnosis.

“Reducing misdiagnosis in the future ‘Digital India’ will remain a challenge without a national body or movement to coordinate patient safety initiatives related to misdiagnosis and strengthening the overall health system.”

Reducing misdiagnosis in the future ‘Digital India’ will remain a challenge without a national body or movement to coordinate patient safety initiatives related to misdiagnosis and strengthening the overall health system. Amid antiquated manual record-keeping systems, there is limited access to diagnostic testing resources and a severe paucity of qualified primary-care providers and specialists. Layering technology on top of these problems won’t produce better patient outcomes unless we address the underlying fundamental problems related to workflow and processes.

Indian medical training traditionally lays greater emphasis on basic clinical examination and history-taking skills, but this might change with digitisation. To maintain physician skill levels, we must use technology to create a learning system where we learn from our mistakes. For example, a general lack of insight into their own diagnosis errors and overconfidence is common among physicians. Currently, there is an absence of effective feedback mechanisms on diagnosis-related performance from hospitals, colleagues or patients. We don’t have a culture of transparency and learning to reduce misdiagnosis, an area in which even U.S. hospitals lag behind. Technology can make things more transparent as long we are prepared to handle this transparency.

The path forward

So what else is needed in the path forward to reduce misdiagnosis in ‘Digital India’? As measurement is the first step to improvement, we must raise awareness of this problem and support health systems research to measure and understand misdiagnosis. We could leverage technology to enable better measurement if we build a culture of transparency and learning. This groundwork could help develop effective strategies to address doctors’ thinking and behaviours as well as address health system flaws, some of which we already know. We could prioritise high-risk areas that can be improved with relatively minor investments, including strengthening primary and emergency care systems. We will need a lot more than 4.1 per cent of GDP to be spent on health care to do that.

The vision of ‘Digital India’ by the current government is inspiring but in a resource-starved setting, setting up a robust and reliable health information technology can be challenging. Policy agenda should first focus on developing and implementing safe and reliable information technology that works all the time. Second, India must develop the workforce and clinical practices required to use this technology correctly and completely. Other countries offer several examples of successes and failures in health IT; we must learn from them rather than reinvent the wheel. Finally, India must leverage technology to improve safety of the health-care system, rather than allowing it to become a distraction, as it happened in Dallas. Multi-disciplinary teamwork is needed with health-care professionals foregoing turf battles and agreeing to work along with experts from policy, IT and patient safety. With a new government working on a new national health policy and a health innovation fund, it must recognise the role low-cost health IT innovations could play in improving diagnostic accuracy, including many that would be useful for rural India.

Misdiagnosis is likely to be one of the bigger health-care safety challenges facing India and solutions are not simple or obvious. While resource-rich nations are still evaluating how to reduce misdiagnosis, we need to start the conversation here and prepare doctors and the health-care policymakers of tomorrow. As we have learnt, even a single misdiagnosis — such as the one in Dallas — can have widespread public health consequences.

(Divvy K. Upadhyay is with the Urban Institute in Washington DC; Dean F. Sittig is with the University of Texas School of Biomedical Informatics, Houston, and Hardeep Singh is Chief, Health Policy, Quality & Informatics Program, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston.)

Printable version | Nov 24, 2014 9:17:50 AM | http://www.thehindu.com/opinion/op-ed/diagnosis-in-digital-india/article6627424.ece

© The Hindu