How doctors think: Lumping vs splitting

Lumping and splitting. On the one hand one, overarching diagnosis explains the whole case and on the other hand the case if finely dissected to include many components. Doctors can err in either direction. In medical school and later in residency -- many eons ago, I learned mostly the splitting approach.  For example, a patient recently presented to my office for a second opinion.   This 36-year-old female complained of an illness which started 2 years ago. First, she noticed fatigue and low-grade fevers. She developed drenching night sweats. Then she experienced diffuse, migratory joint pain and some swelling in both hands. She experienced severe pain on the bottom of her feet making it difficult to get out of bed in the morning and bear weight. She began to experience crushing fatigue-- even showering was a chore and trouble sleeping. She began to experience frequent, pounding headaches which could last for a day or 2 every other week. She experienced brain fog, memory loss, occasional confusion and trouble thinking clearly at times.  She experienced some eye pain and redness along with severe light sensitivity.  She experienced numbness and tingling and a loss of balance. She experienced depression, anxiety and bouts of rage. She had complained to her primary care doctor over several visits. He seemed to blow her off.  The she saw a nurse practitioner who ordered a Lyme test, a Western Blot sent to LabCorp.  The test was read as positive, IgM bands 23 and 41 present.  She was referred to an infectious disease specialist who took a history and looked at the labs. He proceeded to draw confusing graphs-- scribbled on exam table paper, and explained to her why she did not and could not have Lyme and the test was a false positive.  This patient lives in a Lyme endemic area and spends a lot of time outdoors. Many of her neighbors have suffered with the illness.  She has no recollection of a tick bite or a rash. She did some research and thought the ID doctor might be wrong. 

The splitting approach parses out symptoms/problems   A problem list could look like this:  Fatigue. Change in mental status. Eye pain.  Joint pain.  Headaches.  Positive Lyme IgM.  And so on. 

Ddx– differential diagnoses are then constructed for each problem.  For example – this is the non-Lyme doctor’s thinking. 

Fatigue -- Rule out:  sleep disorder, sleep apnea, infection, autoimmune disease, chronic fatigue syndrome, fibromyalgia and others. 

Mental status change or altered mental status:  R/O (rule out) meningitis, encephalitis, brain tumor, intracranial bleed – like subdural hematoma, metabolic/toxin issue – like liver failure and others. 

Joint pain:  R/O rheumatoid arthritis, other collagen vascular illness, crystal arthropathy—like gout, infection, Lyme disease and others. 

Headaches: R/O migraine, migraine variant, tension headache, brain tumor, sinusitis, TMJ etc. 

Peripheral neuropathy:  R/O diabetes/pre-diabetes, thyroid disease, b12 deficiency, multiple myeloma and other cancer – Lyme somewhere down the list.

Positive Lyme test: False positive according to specialist. R/O a true positive. 

Others problems may be deconstructed in a similar fashion.  

The lumping approaches seeks an overarching hypothesis that connects all the dots. An LLMD might diagnose Lyme with a high level of certainty.  A mainstream doctor might think fibromyalgia.

Lumping is a good start but this shortcut may lead to misdiagnosis. For example, complaints include headache (new onset severe headache) and altered mental status. We must consider the possibility of glioblastoma/brain tumor and other serious intracranial pathology. The MRI cannot be skipped.  A non-contrast study is without risk. In addition, the MRI may reveal findings characteristic of Lyme disease. 

I think it is very likely this patient has Lyme and coinfections (Babesia and Bartonella).  Some splitting is necessary. Lyme therapy can be started. At the same time various key studies are ordered. For example:  Sleep study PSM, to exclude sleep apnea and other sleep disorders; blood glucose (maybe A1c), B12 and folate, thyroid screen, celiac screen, immunofixation RE multiple myeloma and MUGUS, selected other tests to excludes other causes of neuropathy. An EMG/NCV might be ordered to exclude CIDP. (Not a definitive test).  Autoimmune disorder may be considered. Testing for rheumatoid arthritis, lupus and select others may offer alternative diagnosis explaining joint pain, eye symptoms and others.  Too much splitting can be problematic and leave the primary problem(s) unattended to.  I saw a patient today who was informed by a previous doctor that Lyme couldn’t be treated with sleep apnea unresolved. Not so. Other patients may have inappropriate delays of therapy because of genetic issues, for example MHTFR epitopes. 60% of the general population has mutations and variants of the gene. 

Splitting leads to a wide rage of considerations interfacing with internal medicine and multiple subspecialties.  Clumping focuses on patterns and connectivity.  The logic is:  It is much more likely that a previously healthy young woman has one diagnosis rather than many – logic dictates the overarching diagnosis.  (Occam’s Razor). A good rule, but not entirely dependable.  Dr. Afrin wrote the book “Never Bet Against Occam.” Using the same principal, his overarching diagnosis would be mast cell activation disorder. The two diagnoses may co-exist in this patient. 

A disease oriented view of medicine – something presented at typical medical conferences, lends itself to a top down approach – lumping.  This approach may lead to rigid, dogmatic views such as those held by the IDSA.  A "Grand Rounds" model, found at some institutions, shows a bottom up approach.  However, where institutions (virtually everywhere) have banned Lyme as a consideration, the presentations, thought to be thorough and complete, are sorely lacking.