by Rick Lee
March 2013 – Evidence-based medicine will always have a place in our health care industry. For example, we can ill afford to unleash unproven drugs on an unsuspecting and vulnerable population. Having the FDA require strict adherence to time-tested processes and methodologies protects our citizenry. But health care software development needn’t be required to adhere to the same strict processes that the pharmaceutical industry must follow.
Imagine Weight Watchers communicating a data point from its relationship with one million clients. Let’s say the company declared that those who weighed themselves daily had a 14% better weight improvement than those who didn’t. Would that observational data matter at all to medical scientists? Would health care publications ignore it as not worthy for publication? Would it make a difference in some of the new business models?
How about if a fitness company reported success in hypertension management related to water consumption, or for those obtaining more than seven hours of sleep on average? Is observational data considered specious by medical scientists and health care experts?
Both the aforementioned examples represent how commercial enterprises improve their business models. Health care companies operating in the bare-knuckled world of high finance, rarely rely on clinical research that comports with all of the highly structured and codified rules of scientific studies. Rather, these companies have embraced the value of rapid iteration or as software experts call it, A/B Testing. Try something quickly. Then try another approach. Compare and contrast the impact of each. Go with the more effective. Then, queue up another A/B test.
So…what happens when health care meets the software industry? Clients in health care clamor for scientific studies that reveal outcomes achieved while adhering to the rigorous standards of clinical research. Software executives counter with the insights gained from rapid iteration. And like two ships passing in the night, the interchange misses the mark.
Let me state my position unequivocally: commerce moves too swiftly to tolerate the ponderous pace of clinical research. Further, customers with hundreds or thousands of participants can ill afford to deprive half of their population from benefitting, because clinical research insists on a control group that doesn’t share the value of an operational breakthrough.
So what is a software entrepreneur, developing code for a digital medical record to do? Experiment, test, observe and repeat. For instance, a great deal of iteration has occurred with creating the right Health Risk Appraisal (HRA). In the mental health field where I first toiled, we were bound to the Minnesota Multiphasic Personality Inventory, a 567-question survey first created in 1939 that was given to all psychiatric patients in our hospital. What seemed obvious to me and most of my colleagues was the challenge of answering the sheer magnitude of questions asked. Consequently, a more recent version, the MMPI-2, has ONLY 370 questions. There are many HRAs that suffer from the same problem…too many questions. What’s the right number?
The right number is the test size that is finished by most survey respondents in under 10 minutes. So start with 100. Reduce to 80. Drop to 60. Shrink to 40. And settle on 34. That would be A/B testing shrewdly applied in a pure, objective approach. See how many people complete the HRA when it is 100 questions. Then keep reducing the size of the appraisal, until you reach the percentage completion that is desired.
In the end, pre/post outcomes that observe the variance from the beginning of the observational period to the end are more than enough conclusive evidence to justify software enhancements to health care solutions. Finding out the right number of questions for your initial HRA or any other metric that requires some form of scientific methodology to underscore your business case, without necessarily hamstringing your enterprise, can be the right way to build a health care business. Paying homage to evidence-based medicine, which slows down progress and incapacitates nimble, swift-acting companies undermines our ultimate goal of fixing health care and making it more affordable.
About the Author
Rick Lee has made a career in the creation and sales of specialty health care applications sold to employers and health plans. He is the CEO of Healthrageous, a behavioral change company launched in 2010 by Partners HealthCare and the Center for Connected Health.
Healthrageous has grown since Series A funding in May, 2010 to serve 16 employers and 4 BCBS plans. It also has a multi-million dollar engagement with Boehringer Ingelheim. The company has 37 employees and is located in Boston’s financial district.
Mr. Lee was the President of Employer Solutions, a $125 million Employee Assistance division of Magellan Health Services (NYSE:MGLN) from 2006-2008.
He founded Quality Oncology, a cancer disease management firm, in 1995 and was instrumental in the company’s growth and ultimate sale to Matria Healthcare in 2003. He was part of the founding team of Value Health, Inc. a specialty managed care company that grew to a $1.5 billion IPO in 3 ½ years.
Rick is a graduate of Harvard College and the University of Washington School of Public Health, where he received an M.H.A.
You can reach Rick Lee at (251)706-4198.