Monte Minute‎ > ‎Monte Minute‎ > ‎

Septic Arthritis, Likelihood Ratios and Acute Kidney Injury - 8/14/2012

posted Aug 14, 2012, 1:38 PM by Rohit Das   [ updated Dec 27, 2012, 7:12 AM by Purnema Madahar ]

Arthritis, i.e. inflammation of the a joint, is a sign we encounter very frequently. It has a very wide differential, and clinically we try to base our deductive reasoning on the distribution of joints involved, the pattern of their involvement, and other associated signs and symptoms. Septic arthritis, which was the focus of our resident report today with Dr. Conigliaro, is the most dangerous of the potential possibilities, and deserves special attention. Lucy Torres presented a case of a 49 year old male presenting with arthritis of multiple joints over a 3 day presentation, and was ultimately diagnosed with polyarticular septic arthritis, in the setting of Group C Streptococcal Bacteremia.

Septic Arthritis is bad news. Even worse, it’s fairly common bad news.

  • In the Rational Clinical Exam (RCE) study attached (a series containing a little over 6,000 patients), 10% of patients presenting with a swollen joint(s) were diagnosed with septic arthritis.
  • Despite improvements in surgical management, the prognosis has not changed dramatically. One-third of patients have a joint-related morbidity, and depending on co-morbidities, there is an associated 7-15% mortality.
  • Polyarticular septic disease (15-20% of cases) has a much worse prognosis, with studies citing a 20-30% mortality, and even worse with Staph. Aureus (50%). That, frankly, is crazy.

Our case, specifically, is a bit odd. The majority of septic arthritis is caused by Staph and Strep, and of Strep organisms – A,B, and even Pneumoniae (especially in polyarticular disease) are more common that Group C Strep...a source for bacteremia was never found for Lucy's patient.

So how do we diagnose this DDD (…Devastatingly Deadly Disease)? Well, if we look at the likelihood ratios cited in the RCE series:

  • Other than having a prosthesis +/- an overlying skin infection (combined, a +LR of 15), history, physical exam, and basic labs are not particularly helpful.
  • So what does it come down to – tap it! The fluid WBC count is the most useful and powerful tool we have, and at a level of >100k, the +LR is around 30 (a change in pre-test probability of 60-70%!!).
  • Bottom line – arthrocentesis is imperative in diagnosing a very dangerous, relatively common disease in patient presenting with arthritis.

Oh, and If all this likelihood ratio talk seems like pish-posh, a good review article is attached. If you read anything that I attach on these dailys, read that…understanding the predictive performance of diagnostic tests = understanding the science behind clinical reasoning – LEARN it, LOVE it.

At intern report, Acute Renal Failure (or AKI, or whatever kidney gurus call it these days) was the topic of discussion. Aman Shah presented a case of an elderly woman, presenting with “failure to thrive,” who had a Creatinine of 12.9 on admission (last Creatinine, in 5/2011, was 0.9). Just to reinforce important concepts:

  • The definition of acute renal failure has been redefined about a billion times. Simplified, within 48 hours:
    • Increase in Creatinine of ≥ 0.3 mg/dL
    • Percentage increase in Creatinine of ≥ 50%
    • Oliguria - <0.5cc/kf urine output for more than six hours
  • In terms of epidemiology, in one series of ARF in tertiary care centers (attached), 45% of cases were ATN, 21% “Pre-renal,” and 10% were BPH. The “cool” diagnoses – GN, AIN, etc., represented a measly 6-7%...ahhh pre-test probabilities...
  • Since ~70% of AKI is due to ATN or pre-renal failure, it’s nice to differentiate between the two – this is where our friend FeNa comes in. Though a value of >2% usually indicates ATN, a value of <1% indicates volume depletion accurately only in patients with severe renal failure/oliguria.
    • In patients with ATN that isn’t too severe, there still may be enough nephrons functioning and reabsorbing Na à low FeNa à false positive for “pre-renal failure”
    • Other things can cause a low FeNa – GN, nonoliguric AIN, and pigment induced tubular damage
    • FeNa is not interpretable in the setting of diuretics, in which case FeUrea becomes a better test

Hmmm, I think that’s about it for today. Couple of things:

1)     Please note Dr. Epstein’s correction regarding PTH's function in the kidney, from yesterday’s daily, if you haven’t already

2)   Please check out the attached pic – showing what happens to 2% of people on chronic bisphosphonates…Courtesy of Dan Massera

Until tomorrow….





Polyarticular Septic Arthritis
Christodoulou et. al., BMJ 2006 Volume 333(25): 1107-1108

Margaretten et. al., JAMA Rational Clinical Exam Volume 297(13): 1478-88

McGee, JGIM 2002 Volume 17: 647-650

Liano et. al., Kidney Int 1996, Volume 50: 811-18

Abuelo, NEJM 2007, Volume 357(8): 797-805

Courtesy of Dan Massera (Thanks Dan!)
Picture of Jaw Osteonecrosis - NEJM Images in Clinical Medicine, August 2012