Rare Diseases in Primary Care

Rare Diseases in Primary Care

Print This Post Print This Post


I volunteer as medical director of an interfaith free clinic in our community. Our staff consists of adult nurse practitioners, a physician assistant, and volunteer family physicians in addition to a physical therapist and behavioral counselors. Recently in our clinic practice we discovered that one of our patients appears to have a rare disease. Here’s how we found out:

Since we are available only one day a week, we recommend to our patients that if they have an urgent problem to call 911 or go to the hospital emergency room where they will be seen regardless of their ability to pay. This patient had an episode of upper abdominal pain and when the ED physician got an abdominal CT, an astute radiologist noticed cystic changes in the base of the lungs and recommended further evaluation.  Some members of the Northern Virginia Medical Society donate services to the free clinics in the area, so I was able to get a consultation from a pulmonologist who thought the findings most consistent with pulmonary lymphangiomyelomatosis (LAM) and recommended referral to a thoracic surgeon for open lung biopsy. Fairfax Radiology Associates donates free imaging to our clinic so I was able to get a full high resolution chest CT which further supported this diagnosis. To protect personal information, I will spare details and give only the bottom line; I had never even heard of this disease, much less considered it in a differential! Fortunately, we are geographically close to the National Institutes of Health in Bethesda, MD where I found an ongoing research protocol for pulmonary LAM. We have referred the patient to the NIH Clinical Center for further evaluation and possible experimental therapy.

According to the National Organization of Rare Diseases (NORD), a rare disease is one that affects fewer than 200,000 Americans at any given time. According to the National Institutes of Health, there are between 6,000 and 7,000 rare diseases affecting from 25 to 30 million Americans. Approximately 50% of the people affected by rare diseases are children.  The NORD website has physician guides to many of these diseases aimed at primary care practitioners (http://nordphysicianguides.org/).

In a 2004 article titled “Zebras on the Commons; Rare Conditions in Family Practice”, Bill Phillips (1) did a study of the rare conditions observed in the practice of 4 community-based family physicians in private practice. He reviewed their office records and found 100 patients with conditions that were either rare or rare presentations of common diseases.  His results were as follows:

“These FPs cared for patients with a wide variety of rare disorders across the spectrum of patient age and sex, organ system involved, and medical specialty area. FPs identified the problem in 89%, diagnosed the disorder in 54%, provided acute care in 56%, and provided continuing care for 76% of patients. FPs consulted other physicians in 85% of cases. The condition was life threatening in 58% of patients.”

In the decade since that survey, the landscape of diagnostic entities, possible findings, and known relations between them has exploded due to primarily to the growth of diagnostic imaging techniques, genomics, and new therapeutic options for diagnoses that were previously untreatable so diagnosis had no urgency.  The replacement of ICD9 with ICD10 has increased the number of diagnostic entities with distinct codes five-fold (http://www.physicianspractice.com/icd-10/six-key-differences-between-icd-9-and-icd-10).  Radiologists are under increasing medicolegal pressure to communicate findings of possible pathologic significance and these reports are often shared directly with patients through the electronic medical record; we are now overwhelmed in primary care with the opportunity (and obligation) to further investigate findings that may never have been captured in previous decades. (http://appliedradiology.com/articles/hobsons-choice-the-new-acr-practice-guideline-for-communication). A similar growth of laboratory findings generated by immunologic and molecular technologies has occurred. (http://www.nap.edu/read/9997/chapter/5#63). All this pales before the potential impact of genomic sequences known to be associated with human disease; there are over 200,000 records currently in the ClinVar database of the National Center for Biotechnology Information at the National Library of Medicine of NIH. (http://www.ncbi.nlm.nih.gov/clinvar/intro/)

All of us in primary care see patients daily who have unexplained findings which may or may not be a threat to their health. One of the major challenges in primary care practice is that because “most hoofbeats are horses, not zebras” we need to be extraordinarily adept at using diagnostic resources most effectively; diagnostic tests which have high accuracy when the prevalence of a disease is reasonably high in the population tested may be almost worthless if the frequency of the disease in the population being tested is low. This is universally true for “rare diseases”. Furthermore, patients may incur substantial expense and risk if the tests are invasive or if the results are either false negative or false positive.

Widespread adoption of EMR technology has been assumed to be a major boon to assist clinicians and patients in diagnostic decision support, however there is currently no such system in common use. There have been many attempts over the years to develop computer-assisted tools for general medical diagnosis. A website http://www.openclinical.org/home.html  chronicles and reviews many of these efforts; none are in widespread use primarily because they all depend on the user recording positive and negative findings.  One that I have found helpful is DxPlain (http://www.mghlcs.org/projects/dxplain/); DxPlain generates a list of common and rare diseases associated with one or more clinical findings. Unfortunately, DxPlain is not currently available outside of academic medical centers; if you are affiliated with one, contact your administration and you may be able to use the internet version of DxPlain with their account. The IBM Watson project promises to automate medical diagnosis, but I have yet to see a comparison with either human diagnostic performance or any other automated system published in a peer review scientific source. (http://www.medicaldaily.com/5-ways-ibm-watson-changing-health-care-diagnosing-disease-treating-it-364394) Unless and until computer-assisted diagnostic tools become routine in primary care practice, I believe it is essential to explain these challenges to our patients and give them a quick lesson in the diagnostic process, the statistical nature of uncertainty, and the limitation of applying “diagnostic” tests designed for binary decisions in specialty care to undifferentiated findings in a primary care population.

There is a free educational resource guide for physicians and nurses managing patients with rare diseases at http://www.raredr.com/resource-guide. It includes patient education resources, a listing of FDA-approved orphan drugs by disease, and a listing of patient advocacy groups by disease.

Charles Sneiderman, MD, PhD, DABFP
Medical Director, Culmore Clinic
Falls Church, VA



  1. Phillips, WR. Zebras on the Commons: Rare Conditions in Family Practice. J Am Board Fam Med July 1, 2004 vol. 17 no. 4 283-286. http://www.jabfm.org/content/17/4/283.long


Published October 7, 2016