Before delving into the issues I mentioned above, let's use a more well known example of these "second order effects." During the toughening of drug sentences in the late 1980s the Anti-Drug Abuse Act made prison sentences proportional to the quantity of illicit substance that a dealer was holding when arrested. The thinking behind such laws is straightforward: stiffer sentences for bigger time drug dealers. At first blush this sounds like a solid enforcement strategy; however, both in theory and in practice these laws had the unintended consequence of incentivizing dealers to hold smaller quantities while still being able to meet demand. Dealers responded to this by placing a new premium on purity. If they could sell their customers half the weight for the same price, they faced a lighter sentence if/when they were arrested. In this way drug laws that were designed to curtail large scale drug dealing had the perverse effect of increasing drug purity, which itself has many second order consequences (higher overdose rates, increased addiction potential, etc).
These negative results to well intentioned laws, policies, and practices are often overlooked but incredibly important. The much maligned WTUs serve as an excellent example of how a well intentioned (well designed, even) system can be victimized by other system-wide issues. Created in the wake of the Walter Reed scandal in 2007 they were intended to be units where soldiers could convalesce before returning to their units without the bureacracy of medical holding units. When the Pentagon initially set these units up they included policies for expediting the replacement of soldiers sent to WTUs. Here we need to pause and revisit the environment all of this was occurring in: it was arguably the lowest point in the public perception of the Iraq and Afghanistan operations and the Army was having trouble meeting its recruiting goals, so standards were slipping somewhat.
To put these various pieces more concisely, you had an overstretched military, a burgeoning problem with un(der)-qualified soldiers, and a new program that allowed for the rapid replacement of a soldier if they could be medically dispatched to a WTU. Commanders planning for an upcoming deployment had few options for their truly problematic soldiers (along with their actually injured); they could begin proceedings to medically discharge a soldier (a bureaucratic nightmare that takes quite some time) or they could transfer them to a WTU and receive an able replacement in time for deployment. Thus commanders were acting in their own (and the mission's) best interests by filling the WTUs with their drug addicts and malingerers. The consequences of this are well documented. WTUs became toxic environments ill-suited to actual recovery by soldiers who want to return their units. The situation got so bad that many Purple Heart recipients were resentful of being sent to WTUs.
News stories about this, detailing the troubles and glimmers of hope belie an understanding of the systemic causes, instead focusing solely on the first-order issues, those that are immediately attachable to the WTU itself. While it is absolutely true that the first-order problems need to be solved. For instance, the staffs of these units need to be adequate in numbers as well as training. More importantly (in my view) is that unless, and until, the time consuming (both in length and officer/NCO attention) process of discharging unfit soldiers is remedied any fix is temporary.
It must be made clear that if we attempt to fix the problem of commanders "dumping" malingerers into these units by changing WTU-centric policies it will merely punt the discharge problems to the future. The issue will lie dormant until another next well intentioned policy accidentally allows an end-run around the problem. The unfortunate collision of well-intention WTUs and the reality of medical discharges has shown an underlying, systemic problem that needs addressing.
The polypharmacy issue among soldiers is similar, except that the systemic problems it reveals are much wider and deeper. The reason there is a problem with polypharmacy among civilians and soldiers alike is the continued balkanization of medicine as a practice. To digress for a moment, in early 2008 the NIH began the Undiagnosed Diseases Program which does precisely what you would intuit from its name. It brings in patients from around the country who have had exhaustive medical workups and run into dead ends. These are not cases the family doctor couldn't handle, these are cases that no doctor in a metroplex could figure out. (As an aside, the program recently "discovered" its first novel disease.) The NYT Magazine ran a piece about the program back in 2009 about a woman there as one of their first patients. What has always stuck out from the piece is this passage:
At least 16 specialists had traipsed past Stiers's bedside in the previous four days; almost all of them managed to carve out time for the spontaneous meeting. Not surprisingly, the dermatologist suggested a dermatologic diagnosis, Vogt-Koyanagi-Harada syndrome. The nephrologist, James Balow, still liked his kidney diagnoses. The rheumatologist didn't want to rule out Sjogren syndrome just yet. "Any other specialists around here who want to explain your own favorite organ?" [Dr.] Gahl asked with a little chuckle.Dr. Gahl's joke is not really a joke at all, it truly is the way that doctors are forced to think. A myriad of issues contribute to this, but to put it succinctly: there is simply too much information for doctors to know everything. As a relavent example, the variety of drugs has grown so large that medical students are taught the mechanisms for particular classes and within that class individual drugs are generally distinguished only by one or two characteristics unique to them. Obviously this is an organized way to teach the information, but it also serves to partition the drugs into discrete boxes that are only occasionally linked by interactions and contraindications.
Simultaneously doctors are taught that drugs treat discrete pathologies and older less-selective drugs are inferior, they're even referred to as "dirty." Indeed, aspirin, one of the most widely used drugs in the world, would not be approved today. The combined effect of this is to teach and constantly reenforce the view that diseases and drugs are akin to bulls eyes on a firing range. Different bullets can be used for each target in order to fix each problem individually. The reality is more akin to the joints on a bridge truss. Individually targetting one or two may have no impact on the integrity of the bridge, but disrupting too many can be catastrophic.
Any efforts at solving the polypharmacy issue that do not acknowledge this underlying, structural issue are doomed to failure. Patients do not know enough to choose which drugs are important and doctors are trained to have tunnel vision regarding the pathology in their area and in front of them.
Often times these second order effects are unpredictable or paradoxical. That they occur should not dissuade us from attempting changes and fixes though. What we need to learn is the need for iterative change. After a new policy or procedure is implemented we must be attentive to what structural issues it reveals, not only in itself, but in the now augmented antecedent as well. Failing to recognize these problems results in makeshift solutions that merely postpone its reemergence. In the words of Mr. Henry, my 5th grade teacher, "There's always time to do it right the first time." In the case of more intractable problems, like the balkanization of medicine, often the acknowledgement of the problem in the first place is helpful in avoiding the attendant pitfalls.