Simplicity and Sharing

By C. Matthew Hawkins, MD

Recently, the oft cited and brilliantly articulate Atul Gawande described the enviable quality and efficiency achieved by the Cheesecake Factory restaurant chain as a possible role model for our health care delivery system.[1] His arguments seem clear. But it left some lingering thoughts, specifically related to the goals of informatics.

Much of the purchasing efficiency that has been achieved by the Cheesecake Factory and other chain restaurant establishments has been secondary to simple software solutions. Tracking consumption, equipment cost, customer traffic, and remaining inventory has enabled chain restaurants to achieve remarkable resource-use efficiency. Why haven’t we developed similar systems or tracked similar metrics in healthcare? Why are our radiology departments staffed to the brim with radiologists from 8am to 5pm, only to be left with a scarcity of manpower in the late evening when emergency department volumes tends to peak? Maybe this approach is appropriate, but I’ll wager that many departments do not have the data to back that up.

I’m sure many of us have heard or spoken the following phrases: “I think it’ll be busy tonight since it’s Saturday.” or “I bet it’ll be slow tonight since it’s supposed to snow.” Was it slow? How slow? How many people have sat around staring at their colleagues on a holiday shift? Were you surprised that the ED was not busy on Christmas morning? I wasn’t. It was just like last year and the year before that…but maybe it could have potentially possibly been somewhat marginally busy for a short while. So it’s a good thing we were all there.

If restaurants can precisely predict customer volume and food/drink consumption for NFL playoff game days and January Wednesdays in Sioux Falls, why haven’t we developed similar prediction modeling for our health care delivery systems? Simple metric measurement such as that done in the restaurant industry could allow us to appropriately allocate our physician resources and predict technologist staffing requirements for high and low volume days, hours, and moments. Sadly, the metrics, software, and informatics solutions I speak of are simple. Very simple. We’ve discovered how to quantifiably measure intrahepatic iron content and liver stiffness with MRI, but we haven’t taken the time to track how many CTs we perform on average from 5pm to 9pm on Mondays in February.

Clearly, there are facilities and institutions that have taken initial steps to predict volume and resource consumption in an effort to maximize resource allocation efficiency…no doubt. But this isn’t about who is and who isn’t. Instead, this is a moment to pause and remind ourselves of the inherent value of simplicity. In the field of informatics, it is easy to be lured into a quest to discover the next billion dollar solution. Or develop an esoteric complex software package that only a few providers at the wealthiest institutions can afford. It may be, instead, that the simple solutions — the solutions that seem obvious to us all and that are readily available — may be the tools we in informatics can harness to impact the medical industrial complex on a wider scale.

Lastly, Dr. Gawande’s column re-kindled a burning concern that I have for healthcare and its unavoidable evolution. As our healthcare delivery system continues to encourage (intentionally or unintentionally) the formation of large conglomerates battling in a price-driven, commodity-laden open-market operating on thinner and thinner margins, will we maintain the same willingness to share knowledge throughout the medical community? What would be the benefit? We’re so deeply immersed in our current culture of open knowledge sharing, ego-boosting journal publications, and national meeting presentations that envisioning an information-hoarding medical community that functions in secrecy can be difficult. But, if two multi-billion dollar health systems are fighting for patients in the same metropolitan area and one of the systems can guarantee and prove a lower complication rate and shorter hospital stay for knee replacements, will that system (and its employed physicians) be willing to share their new techniques with the world so that all healthcare systems can offer the same results? If an employed physician at hospital X discovers a new MR sequence that diagnoses labral tears with 100% accuracy, would he/she be allowed to present the science behind their sequence at a national meeting so that their colleagues can benefit? Will they publish their detailed data? Maybe they would. But the skeptic in me thinks otherwise. I haven’t seen Apple and Samsung sharing many ideas lately.

When ACE-inhibitors were indubitably proven to improve outcomes in patients with congestive heart failure it took nearly 18 years for their use to propagate throughout the medical community accordingly. 18 years. That was in a culture of knowledge sharing. Imagine if the ability or willingness to share knowledge is crushed by health systems competing in an open-market?

What kind of culture do we want to establish in the field of informatics? Are we going to be willing to share or eager to hoard? This is not an “open-source versus proprietary” question. This is, perhaps, bigger. When your informatics team stumbles upon the next great technological innovation in healthcare delivery or informatics solution that gives your radiology department a leg-up on the competition, will you share the idea with the guy across town? Right now, the answer seems obvious…or does it?