Florence Nightingale: Healthcare Analytics – by Roopam

Healthcare Analytics

In 2012, Larry Smarr concludes his TEDMED talk (a TED conference for medicine, healthcare, and biology) with the phrase:

‘Because of big data and because of our ability to analyze it – we got hope’

Larry is no medical professional or biologist rather he is a physicist (astrophysicist), and computer scientist by profession. Before we learn more about Larry Smarr and his path-breaking work in healthcare analytics let’s look at a page from the history of statistics and data analysis in medicine and healthcare.

The Lady with the Lamp

In 1853, Crimea war broke between Russia and an alliance of European countries. Florence Nightingale had volunteered to nurse British soldiers fighting for the alliance. Her nursing effort to save hundreds of lives won her a special place in world history. However, her lesser known work with statistics and number crunching to analyze mortality data initiated sanitation reforms in healthcare that possibly saved millions of lives. When she arrived at the military hospital in Scutari during the war of Crimea, she was appalled by the lack of sanitation. She hypothesized that unhygienic conditions in the hospital were a major reason for deaths. In her effort to test her theory, she collected data meticulously – this was an unusual scientific practice in healthcare at that time.

Nightingale’s diagram (click to enlarge)

Nightingale, along with statistician William Farr, devised some of the most innovative & aesthetically pleasing data visualization techniques to analyse data. Through her careful analysis, she confirmed that diseases such as typhoid,  typhus, and cholera caused by unsanitary conditions in hospitals were the primary cause of mortality. This pioneering work of Florence Nightingale highlighted the importance of hygiene in hospitals and hence initiated a whole lot of reforms in healthcare practices focused toward hygiene and cleanliness. Florence Nightingale – also known as the lady with the lamp – showed us the light and the way to improve healthcare with the usage of statistics more than a century and a half ago.

It is a bit sad that despite a long history of successful usage of data and analysis in healthcare, the industry has not made the same progress with big data and analytics as other industries like banking, and retail. There are several reasons for this. But before we discuss those reasons let’s take a time leap of more than 150 years from the times of Florence Nightingale and come back to Larry Smarr and his act of..

Quantifying His Body

In early 2000, 51-year-old Larry Smarr moved to University of California, San Diego as a faculty. At that time he weighed more than 200 pounds. All the lean and fit people around him in California made him seriously consider reducing his weight. Smarr being a physicist was well trained in the scientific processes of measurement, and analysis to decipher natural systems. He considered his body no different than any other natural systems. In his quest for a fitter body, Smarr like a true scientist started measuring every possible parameter in his body to enhance his fitness. His skills in computer and data science enabled him to analyse this data. Through this process, he self-diagnosed himself with many medical conditions and prepared himself well for the to be followed diseases.

Healthcare analytics: Larry Smarr measuring himself

In one of his early analyses of the data, he noticed an anomaly in one of the parameters linked to inflammation in his body. He noticed it was fluctuating 10 to 15 times above the normal range despite his well-controlled diet.  He visited his doctor (several doctors actually) with his findings. The reaction of almost all the doctors to Smarr’s data highlights a specific shortcoming in our medical system: the doctors asked Smarr about his symptoms. Since Smarr had no symptoms of a medical condition but just data, doctors termed his findings as academic with no medical relevance. Several weeks later, Smarr experienced severe pain in his abdomen and was diagnosed with diverticulitis – a disease caused by high level of inflammation in the body.

Doctors are trained to treat patients’ existing symptoms and their training doesn’t equip them to tackle data that suggests future problems. With the advent of big data collection and the power of data mining, these prognostic signs will soon become a norm rather than exceptions. Medical science and healthcare professionals need to prepare for this new era of data-driven healthcare and prognostics.

Later in 2011 Larry Smarr, while working with his bio-markers noticed another anomaly in his white blood cell functioning. He studied these signs in medical literature and narrowed down the signs of Crohn’s disease. His analysis led him to diagnose his disease much ahead of time before the disease started debilitating his body and showed up in the form of symptoms that doctors could treat.

Healthcare Analytics – Data Challenges and Opportunities

The annual cost for Larry Smarr’s kind of self-monitoring is close to $ 10,000. Additionally, there is a one-time cost of expensive self-monitoring devices (from Fitbit, Omron, Body Media etc.), and the additional cost of data monitoring and analysis. This is certainly not affordable for all. However, there is hope. The first effort to sequence human genome cost more than a million dollars and more than 10 years to complete. However, today you could get your DNA sequenced in a couple of thousand dollars with a delivery time of less than a day. There are claims that in the next few years you could get your entire genome sequenced in less than $ 1000. We have seen this exponential decline in cost with improved efficiency in virtually all walks of technology. Healthcare technology won’t be any different. Soon most humans, if not all, will be able to afford this technology like a cell phone.

Electronic Health Record (EHR) was one of the first efforts by the Obama administration for improved healthcare. The idea is to store data for all the hospital & clinical visits by every patient in a usable format similar to customers’ banking transactions. This is not an easy task. The healthcare industry is much more complex and rapidly evolving than any other industry. The routine medical standard operating procedures get obsolete faster than SOPs in any other industry. Despite all the challenges, developing a robust EHR and linking it with other sources of data including clinical trials, genomics, proteomics, and self-monitoring (similar to Larry Smarr efforts) has many unparalleled benefits.

Unified Data for Healthcare Analytics

The foremost benefit of unified data is that since the planet has an acute shortage of well-trained doctors, medical professionals, and medical facilities, a well managed unified data, and healthcare analytics will ensure right and timely healthcare to the patients with the highest need. These patients will be identified with technology supported by artificial intelligence and machine learning. Moreover,  healthcare analytics will push medical science from reactive to proactive care as we have learned from Larry Smarr’s self-diagnosis of his Crohn’s disease. Healthcare analytics essentially will make the patients better equipped to tackle his or her future ailments.

Sign-off Note

The lady with the lamp, Florence Nightingale, showed the light and initiated a transformation in healthcare that saved millions of lives. Healthcare analytics, in my opinion, holds a flashlight that will take us to the next era of healthcare and medicine.