Delayed Dx in Patient With Persistent Back Pain, Diarrhea

A 21-year-old man presented to a hospital upon referral by an outside provider for assessment after 7 weeks of intractable left back pain. He had spent the previous 3 weeks hospitalized at another facility for intense back pain. During this time, he developed fever and pyuria; urinalysis indicated a susceptible Escherichia coli infection, which was treated with doripenem for 2 weeks. His fever abated, but his back pain persisted.

Although his clinical condition improved to some extent, laboratory tests performed at the referring institution revealed that his serum albumin level was 1.8 g/dL (normal range 4.1 to 5.1 g/dL) and his C-reactive protein (CRP) level was 7.8 mg/dL (normal range <0.3 mg/dL), which prompted his referral for further investigation.

At his initial assessment, the patient reported an unexplained weight loss of 5 kg and diarrhea up to 5 times daily for the past 2 years. He did not report any joint, abdominal, or muscle pain; loss of appetite; vomiting; rigors; urinary symptoms; blood in his stool; or high-risk sexual history.

On physical examination, clinicians noted that the patient appeared well, and was alert and oriented to time and place. His temperature was 37.0°C, his pulse was 98 beats/minute, his blood pressure was 104/59 mm Hg, his respiration rate was 14 breaths/minute, and his oxygen saturation was 98% on room air.

Findings from lung and heart examinations were unremarkable, and there was no evidence of swelling or skin abnormalities. His abdomen was soft and flat with slight tympanic bowel sounds. Palpation revealed mild tenderness in the right upper abdomen, but there were no signs of muscular defense or rebound tenderness. However, the patient did have left costovertebral joint angle tenderness.

Among the findings from urinalysis, urinary protein level was 1.92 g/g Cre (normal range <0.15 g/g Cre), white blood cell count was 50 cells/high-power field, red blood cell count was 10 cells/high-power field, and bacteria was present.

Blood test results revealed that his CRP level was now 6.3 mg/dL, and his serum albumin level was still 1.8 g/dL. Laboratory testing results showed that the patient's liver function was within normal limits.

Clinicians diagnosed the patient with a urinary tract infection (UTI) with complicated pyelonephritis. In light of his persistent diarrhea and elevated inflammatory response, they suspected that he also had inflammatory bowel disease. To investigate further, they performed a colonoscopy, which revealed numerous longitudinal ulcers and skip lesions. Results of the biopsy suggested a diagnosis of Crohn's disease (CD). The patient then underwent abdominal x-ray and echography to rule out possible post-procedural perforation.

Based on the patient's reported history of pain related to scoliosis, he was prescribed treatment with non-steroidal anti-inflammatory drugs.

Three days after his colonoscopy, the patient presented again to the emergency department, reporting recurrent fever and worsening back pain. Urinalysis on this occasion identified fecaluria, which clinicians suspected was due to a vesicointestinal fistula.

The team re-examined the CT images provided by the referring hospital. These revealed air in the bladder, which clinicians believed also could have been due to a CD-related vesicointestinal fistula; the patient underwent a contrast enema, which confirmed the presence of vesicointestinal fistula.

The patient was admitted for evaluation of the fistula, and started on treatment with tazobactam/piperacillin (4.5 g every 6 hours) and oral mesalazine (3 g/day). Fourteen days later, he underwent laparoscopic resection of the terminal ileum due to stenosis, with spontaneous closure of the fistula facilitated by medical management only.

After an additional 2 weeks under observation, the patient had no postoperative complications and was released from the hospital. During ongoing follow-up assessments every 3 months, there has been no recurrence of UTI.

Discussion

Clinicians presenting this case of a 21-year-old man with a persistent UTI despite 3 weeks of antibiotic treatment noted that the case "presents a clear and interesting example of cognitive biases including satisfaction and anchoring biases, that is, finding one disease, the discovery of which prevents the accurate and timely diagnosis of another."

Although the series of diagnostic errors that delayed diagnosis of the patient's CD began with "the referring physician's lack [of] knowledge about complicated UTI and its epidemiology in young men," the case authors acknowledged their own oversight in failing to assess the patient for possible anatomic causes of the UTI.

They noted that their delay in accessing and interpreting the original imaging studies that showed free air in the bladder caused them to incorrectly anchor their initial diagnosis. Subsequently, the discovery of the patient's history of chronic diarrhea "led to the Crohn's disease diagnosis via colonoscopy, [which] generated our satisfaction bias." This error might be avoided by performing a thorough physical examination, they added.

"We failed to consider the complications of CD, such as fistula formation," they wrote, noting that this turned out to be "the cause of this patient's recurrent UTI."

Finally, that they mistakenly ascribed the patient's lower back pain to reported pre-existing scoliosis without performing a detailed assessment of this symptom led to "delayed emergency care for the now-recognized vesicointestinal fistula," they added.

According to a multicenter retrospective study of CD patients, 75% of entero-urinary fistulas were diagnosed in men.

Furthermore, among all CD patients with urinary tract fistulas from a single-center analysis, "fistulas originated from the ileum (64%), colon (21%), rectum (8%), and multiple sites (7%). Urinary tract sites included bladder (88%), urethra (6%), urachus (3%), ureter (1%), and other (1%)."

The group reported this case to encourage clinicians to consider several aspects of the diagnostic process: the frequency and percent probability of a disease; recognition that a disease may be identified by both its organ-specific pathology and its adverse effects on neighboring organ systems; the importance of focusing on patient concerns through use of the standard review of systems, as opposed to "anchoring on data provided by referral institutions, which may be qualitative and devoid of useful patient-driven information"; and vigilance for their own satisfaction and anchoring biases.

To avoid the influence of bias, the authors advised clinicians to begin with "a careful interview and physical examination, as if the patient were a first-time patient." Finally, they emphasized the importance of making "calm decisions with every patient and every case," and recommended using a "check list or problem list for complicated patient diagnosis and management."

Comment