The anatomy of problem solving

How often do we think about what goes into solving a problem or making a decision? There are always at least half a dozen separate opportunities to take a wrong turn, so it pays to review the components of problem solving and decision making. In many ways, the medical model works well.

1. What is the problem? In medical diagnosis, most problems present as a symptom rather than a diagnosis: dizziness, headache, abdominal pain, itch. In order to treat the illness, it is essential to diagnose it and understand the context and impact on the patient. Similarly in management, equating low revenue with poor provider productivity is likely to result in ineffective or harmful solutions. Failing to meet wRVU targets is a symptom, not a diagnosis. The problem may lie in an inappropriate target, in failed systems that impede rather than support providers, or in demand issues.
2. Who is responsible? In medicine, not all problems are amenable to solution by the PCP. Patient financial distress due to a bad economy is beyond the power of the best clinician to resolve. Inability to fix it does not mean there is no value to understanding the issue and taking it into account when solving other problems, of course. Similarly, many problems in a health care organization are caused (and ‘owned’) by other parts of society: third party payors, governments, or patient behaviors, for example. It amy be necessary for the hospital or provider office to study and understand the requirements for documentation or the limitations of coding, but this is in order to successfully work within the parameters of possibility, not with the intent of changing the rules. 
3. Who are the stakeholders? Successful treatment of a complex medical illness depends on active participation on the part of the patient, the patient’s support system, and often a collection of medical professionals. Thus, the patient with COPD and ischemic heart disease having a knee replacement will need to participate actively in their own pre- and post rehab care, but will also need help from family and friends, cooperation from their employer, and coordinated or collaborative participation from their PCP, orthopedic surgeon, radiologist, anesthesiologist, nursing staff, PT/OT, social worker, billing office and payor, and possibly pulmonologist, cardiologist and hospitalist. Similarly, solving a problem that touches many parts of an organization requires participation from every area involved. (I have written previously about the benefit of problem solving in the public space.)
4. What  resources, information, and skill sets are needed? Surgery requires information about the patient and the diagnosis, as well as a surgeon and anesthesiologist supported by OR staff and equipped with surgical and anesthetic instruments. A revenue problem cannot be solved without various analysis skills using information about all the steps in revenue generation, and a patient workflow issue cannot be resolved without???
5. How quickly does the problem need to be resolved? In medicine some problems need to be addressed instantly, often by standardized protocol without prolonged assessments (respiratory or cardiac arrest), some problems need expeditious response but allow a varying amount of deliberation prior to and during the process of treatment (seizure, hypovolemia, potassium of 6.5 or glucose of 800 with a pH of 7.25), some problems allow significant freedom for data collection and deliberation before action is taken (sore throat, joint swelling and pain), and some problems demand significant time and effort collecting and integrating information, deliberation and collaborative decision making with the patient for successful treatment (chronic migraine, moderate persistent asthma).
6. What is the best process for this problem? Some medical problems are best solved with discrete serial steps in an iterative process, such as getting asthma or diabetes under control, while other problems require a broad-based and multi-prong approach, such as simultaneous fluids resuscitation and electrolyte and acid-base corrections, antibiotics and IV insulin for diabetic ketoacidosis related to an infection. Similarly, adding functionality to an eHR is best done in small and discrete pieces, with training and practice resulting in mastery of each step before moving on, but instituting a check-list process to reduce procedural infections is best done as one intervention.
7. How do we know when the problem has been solved? In medicine, we don’t keep adding blood pressure medications after we reach our target. In some settings, we use a standard ‘dose-duration’ for an intervention, such as antibiotics for an otitis. Both of these require monitoring of selected appropriate metrics to assess the response to our intervention as well as a definition of success. Improving clinical documentation by continuously adding forms or check boxes without regard to the issue at hand or the impact on other systems would be like increasing the insulin dose without regard to fasting sugars or the A1c.

A contemplative pause to review these seven items at the front-end of problem solving would go a long way toward preventing bad outcomes.