Evidence Based Practice is a prerequisite for adoption of Big Data in Healthcare Management

Evidence Based Practice has its origin in clinical practice and it is now being applied to the general management area. But the progress is not very encouraging.

The scope of healthcare administration is very wide and depending on the expertise of the incumbent, it may be extended to some clinical aspects as well apart from HR, facility management and Finance. As of today, these functions are largely managed by Healthcare MBAs with or without a medical degree. It may be mentioned that MBAs are not always welcome by the physician community. As for example, in an article titled: “The Hospital as a Factory and the Physician as an Assembly Line Worker” by Arthur Gale in Missouri Medicine,2016, Jan-Feb issue commented: “We see the hospital as factory and our hospitalist group as an assembly line that is in the business of manufacturing perfect discharge”. He also mentioned, “MBAs tell physicians how to practice medicine. Cost rise and quality drops. Patient, Physicians suffer.”  This is not an exception, if you interact with physicians on this issue, you will most likely find that many from the medical fraternity share the same sentiment. Like it or not MBAs have established the legitimacy of their existence in the healthcare sector. All major B’ Schools around the world do have an MBA in Healthcare courses and may have a dual programme as well.

The job of healthcare administrators is to promote efficiency, patient satisfaction, and financial sustainability. They are also responsible for mitigating employee turnover, reducing hospital errors, accountable to board members and managing within budget.  Brent Sigut (2013) has recognized the conflict between physicians and managers. He mentioned, “The doctor believes that the manager is simply a ‘bean counter’ at his core” and he also mentioned, “Healthcare managers foster contention based on their perception that doctors believe managers’ job function are insignificant and not grounded in evidence-based practice”. Here I question: do our managers are evidence-based in healthcare or in other sectors? There is no evidence that managerial decisions are evidence-based. There may be some exceptions but in general, it is not.

I assume that our physicians have exposure to evidence-based practice. Of course, to what extent they practice may be questioned. In a recent survey, in India, it has been found that 62% of respondents (urologists) feel EBP is an added burden to their practice. Evidence-based management has not yet taken root in our management education system. If both, the doctors and managers communicate, argue, debate based on evidence-based logic the conflict will be reduced and at the same time healthcare cost and quality will be improved. Big Data, if properly used, will generate new evidence and a manager or a doctor will accept and use it only when the mindset is “evidence seeking”.  Of course, there is an issue of the trustworthiness of evidence. You find egg yolk is bad for cholesterol in one day and the next day you find it has no correlation. If you follow evidence that comes in media, either you will take everything or you will stop eating.

In order to bring efficiency in healthcare some institutions are introducing the operational practices used in the manufacturing industries in healthcare. As for example, In India, there are two examples of applying assembly line approach in surgery-one in eye care by Arvind Eye Care System and Naryana Hrudayalaya Hospital in cardiac surgery. And the record says that the quality is also not compromised and cost has come down. In both cases they have followed the principles of assembly-line – division of labour by Adam Smith and Taylor’s principles in designing standard work. These two principles are well evidenced. This has helped to deskill of jobs and promote specializations. Both have also adopted appropriate technology, like telemedicine in extending healthcare in remote locations.  We should be careful about selecting technology. As for example, FDA cautioned, “"There is limited, preliminary evidence that the use of robotically-assisted surgical devices for treatment or prevention of cancers that primarily (breast) or exclusively (cervical) affect women may be associated with diminished long-term survival,".

Evidence must also be used in case of managerial decisions as well. In the United States, more hospitals are going for implementing Business excellence and participating in MBNQa, Lean, Six sigma, etc. But scientific evidence of long-term effectiveness of these approaches is not there in the literature. For example the number of manufacturing units participating in MBNQA has come down to one that too from the SME sector while the scheme was designed primarily for the manufacturing sector. Implementation of these practices takes time, money and a huge amount of man-hour. The same is the case with TPM. Our analysis shows that the performance of TPM award winners is no better than their competitors who have not gone for Award and there is hardly any change of business outcome before and after the award.  When the employees do not find gains in implementing some change, they get frustrated and morale comes down. In such cases, one should go for the evidence-based approach. Evidence should also be sought for implementing Big Data before taking any decision to adopt it. It needs infrastructure and specialized manpower.

In a recent paper, “Surprising power of small data” Lee Simons has mentioned the research of Mohsen Bayati that claims it is better to have a small set of biomarkers rather than a long list of them which confuses and increases the probability of false positive or false negative. Physicians are accustomed to taking clinical decisions based on their own data, experience, and judgment. There are structural issues like an investment in IT infrastructure in our public hospitals are almost negligible. As a result, it is difficult to get standardized consolidated data at the aggregate level from which evidence may be sought. Nevertheless, data are maintained at Silos- integrating it at the institute level is also not easy. It is a case of chicken and egg dilemma- evidence is not there as data are not available and data are not collected as we are not evidence-based. Big data will help develop evidence at less cost. Though RCT is the gold standard for EBM, it is costly and difficult to implement in all cases considering the demographic, socioeconomic conditions and the mere size of the population in India. Big data may help get evidence and improve the quality of the decision. It is widely agreed that data is maintained mainly for administrative and billing purposes. The quality of such data can be questioned before its use for healthcare decisions. We frequently come across a complaint from patients, in media, that though medicines were billed but the medicines have not been used. Therefore, it is doubtful as to how factual data may be obtained. There is an inherent paradox in using Big data, which holds good for population and modern medicine is claiming in future treatment will be individualized, which requires quality small high dimensional data. We have also to remember that while EBP is for generalized treatment, medical care is individual-oriented. A recent paper titled: Big Data Big Problem: A Health Care Perspective highlighted the adverse effect of Big Data on patient outcomes. As for example, Google Flu trends thought to be the most successful big data project. But it overstated the number of units by a factor of 2 in 2013. And in 2015 it was closed.

At this point let me share with you some of the research findings from scientific literature regarding the use of Big data in Healthcare Administration. A study based on interviewing 68 health care administrators in 2006, suggests that there was a low-level attitude towards Evidence-Based Management in the US. But a recent study carried out in 2016 indicates there is a change of attitude. It is more positive towards EBM.77% were interested in EBM training and 83% informed they had not received any training on EBM. A study based on a self-reported  survey, published in the Academy of Management Journal in 2017, suggests that the top three pieces of evidence that had been used are Professional Exp., Organizational Data and stakeholders’ views. Evidence from scientific literature was not mentioned. A large-scale survey, across USA, Europe, and Australia, was conducted by EBM stalwarts like Eric Barneds, Denise Rousseau and others in 2017 with responses from 2789 management practitioners. Findings are quite interesting. Major evidence used are - 91% personal experience,64% intuition, knowledge acquired through formal education (71%). Most managers do not read academic journals (70%) and only 14% read peer-reviewed journals. 63% disagreed with the applicability of research findings as their case is unique. Interestingly 73% feel EBM will improve quality of decision and 62% agreed that EBM should a part of formal education.

Some good news for India. Govt. of India is considering to set up a National eHealth Authority (NeHA) for standardization, storage and exchange of electronic health records of patients as part of the government's Digital India programme. Ayushman Bharat Pradhan Mantri Jan Arogya Yojona is not only benefiting a huge number of people but also generating large amounts of data. It has already revealed two astonishing facts, the number of hysterectomies and cesarean birth went up unusually high. It also indicated that 34% of all tertiary claims are towards cancer care. It also helps in getting a segmented wise number of cancer patients. This will help design required intervention initiatives.

It is possible to take a well-argued decision only when both the doctors and managers refer to the same judge for arbitration- which is Evidence. We have to keep in mind that nothing should undermine the physicians’ ability to take clinical judgment. I can take my case as an example to argue for the case. I was undergone a nuclear imaging test for a cardiac health check-up to see whether any block is there. The report says there is no evidence of any blockage however the physician make appropriate decisions based on clinical findings- a disclaimer or risk-mitigating steps for the radiologist. Another example, a report I read in TOI 3-4 years back. The report said about a new pregnancy test, which claimed 99% accuracy compared to the usual accuracy of about 91%. When a reputed Gynecologist was asked to give his opinion- he said it does not add any value to him as at the end of the examination he had to inform the patient whether the patient is pregnant or not. He could not say that she was 99% pregnant!

It is difficult to convince the managers in general including healthcare administrators to practice EBM, Globally and India is not an exception. EBM or EBP should be a part of the formal management course.

Note: This article  is prepared based on my Key Note address, given in International Conference, EBESCON-2019, at SUMANDEEP VIDYAPEETH, VADODRA, 12^th to 14^th December 2019.