Adding Versus Subtracting
One of the first “big ideas” I encountered when I began my higher education was the notion that people tend to be either lumpers or splitters. Now to be clear, like all such dichotomies, no one is always going to take either position—everyone does some of both. But the idea is that any time one is trying to sort something into classes, there will always be some doubt whether a minor variation constitutes an important distinction warranting creating a separate class. “Lumpers” tend to have a high resistance to subclassifications while “splitters” tend to emphasize the differences and resist grouping. Let me give a specific example to illustrate why this human tendency has clinical importance.
Primary diseases of the kidney are typically grouped into those involving the blood vessels, the drainage system, (interstitium,) and the filters (glomeruli). Primary diseases of the glomeruli are classically grouped by the pattern of injury seen on microscopic examination of a biopsy, but as newer forms of study continually lead to further divisions and reclassifications. Sometimes, but not often, these newer subdivisions are associated with improved therapeutics and better clinical outcomes. By temperament, I am a lumper, a usually reserve new subclassifications for those that need a different therapeutic approach. But I had a younger partner who was very much a splitter, and every atypical feature on either the clinical presentation or the biopsy sample would cause her to do searches in PubMed and agonize over decision making. I want to emphasize that neither approach is perfect. My approach was “efficient” but there is a substantial risk of “jumping to the wrong conclusion,” while her approach was “labored” and ran the risk of “paralysis by analysis.”
This tendency to be a lumper or a splitter may parallel another example of dichotomous thinking. Meyvis and Yoon wrote an editorial recently on a study showing that “adding is favored over subtracting in problem solving.”
“Adams and colleagues analyzed archival data and observed that, when an incoming university president requested suggestions for changes that would allow the university to better serve its students and community, only 11% of the responses involved removing an existing regulation, practice, or program…
Adams et. al., demonstrated that the reason their participants offered so few subtractive solutions is not because they did not recognize the value of those solutions, but because they failed to consider them…It thus seems that people are prone to apply a ‘what can we add here’ heuristic…
We propose that the bias toward additive solutions might be further compounded by the fact that subtractive solutions are also less likely to be appreciated. People might expect to receive less credit for subtractive solutions than for additive ones.
…There are many real-world consequences of failing to consider that situations can be improved rather than adding…On a grander scale, the favoring of additive solutions by individual decision-makers might contribute to problematic societal phenomena, such as the increasing expansion of formal organizations and the near universal, but environmentally unsustainable, quest for economic growth.
Certainly, the preference for additive solutions undergirds the growth of bureaucracy as predicted by C. Northcote Parkinson in his classic study. But it is not only business concerns that have to deal with the bias toward additive solutions. Consider how the medical profession adjusts to new imaging technology. How often is a CT scan done for some problem followed up by an MRI or a PET scan for the same problem? A subtractive solution, deciding which imaging was “best” for which kinds of problem, would save money, hassle, and exposure to ionizing radiation, but we continue to just keep adding. Every new automated therapeutic protocol will add lab work as a “safety” measure, adding lab tests that have low utility to those already being ordered by the physician. (Consider the utility of lab confirmation of high bedside glucose meter readings or hemoglobin measurements in patients on heparin.)
All is not lost. We are not doomed to ever expanding, smothering bureaucracies—we simply have to remind ourselves to consider subtracting things that we “have always done” that no longer contribute. As I used to tell the staff in my dialysis unit, if it does not have something to do with good patient care, safety, or getting paid, we don’t need to be doing it. You won’t win any popularity contests taking that position, but you might be able to make patient care and staff experiences better.
We all need to remember that lumping and subtracting are sometimes better and in other times splitting and adding are better. The real challenge, and probably real wisdom, is knowing which mindset is likely to be more productive in facing particular challenges.
12 July 2021
 Meyvis T, Yoon H. Adding is Favored Over Subtracting in Problem Solving. Nature 7 April 2021. doi: 10.1038/d41586-021-00592-0. Accessed 9 May 2021 at https://www.nature.com/articles/d41586-021-00592-0.
 Parkinson CN. Parkinson’s Law and Other Studies in Administration. (1957).
Barriers to Innovation
Innovation is critical for organizational survival, but internal and external forces make it difficult.
What lessons does building cathedrals have for healthcare reform?
Getting to Why
Reduced government spending by definition means someone's income goes down. To mitigate those changes medical organizations need to get to "why."
Playing it Safe?
Many health care organizations have become frightened—so much change is being forced upon them it seems insane to voluntarily try something different. But what it it is essential?
Wouldn't it be wonderful if we got rid of stupid stuff?