We’ve long been staunch supporters of the CAUTI pay-for-performance (P4P) measure. Avoiding CAUTI’s was one of our very first quality improvement projects on a neurology service way back in the 1980’s. The key to success is obviously avoiding unnecessary use of indwelling urinary catheters and limiting duration in those cases where such originally has a legitimate indication. So we were on board when CMS began its program of nonpayment for CAUTI’s. But now, multiple years into the CMS program, we’re beginning to see that there may be unintended consequences of continuing to use CAUTI as a P4P measure.
In our November 2012 What’s New in the Patient Safety World column “CAUTI Conundrum” we pointed out a recent article (Lee 2012) that concluded that the financial disincentive of CMS’ nonpayment policy for hospital-acquired conditions did not lead to lower rates of CAUTI. Rates for CAUTI’s were already declining prior to CMS’ implementation of its reimbursement policy. Specifically, for CAUTI’s the preimplementation decline was 3.9% per quarter and only 0.9% per quarter afterward.
Though we discussed many potential explanations for this, we think that it simply reflects that we were already targeting CAUTI’s long before CMS implemented their policy so there was little change in CAUTI rates before and after implementation.
But in that column we cautioned that when we implement “evidence-based” quality pay-for-performance metrics we sometimes find out later that either the evidence was not so strong or that we get unintended consequences. We think that the evidence base for preventing CAUTI’s is strong. But what about the unintended consequences? We noted that it would be interesting to see how many patients got unnecessary antibiotics for asymptomatic bacteruria because of the CMS nonpayment policy for CAUTI’s. If the latter rate is high that might be a reason to change the metric.
The U.S. Preventive Services Task Force Guideline on Screening for Asymptomatic Bacteriuria in Adults (USPSTF 2008) recommends against screening in men and nonpregnant women. The USPSTF does recommend screening for asymptomatic bacteruria in pregnant women at 12-16 weeks’ gestation or at the first prenatal visit, if later. Other indications include those undergoing invasive urological procedures where mucosal bleeding is likely and possibly surgical procedures with implant material (Wagenlehner 2005, Nicolle 2005).
Even though treatment of asymptomatic bacteruria is not recommended in most circumstances we have seen multiple hospitals in which urine cultures are done on admission so that if the patient later develops a UTI following an indwelling urinary catheter the hospital can say this was “present on admission” and not have to report it as a CAUTI. Some of those patients do get antibiotic treatment for their asymptomatic bacteruria. And while there is little evidence of a beneficial effect of treatment in such cases, the treatment has the potential to have adverse effects.
The Infectious Diseases Society of America (IDSA) guidelines note that potential unintended consequences of treatment of asymptomatic bacteriuria include subsequent antimicrobial resistance, adverse drug effects, and cost (Nicolle 2005). Other potential unintended consequences would be increased hospital lengths of stay, delays in procedures, and the potential for diarrhea and C. difficile infections.
Now a new study (Drekonja 2012) makes us further question whether the CAUTI measure (as either a patient safety measure or P4P measure) may be inadvertently leading to other unintended and unwanted complications. The authors did a retrospective review of urine culture practices preoperatively in patients undergoing nonurologic surgeries in a VA hospital setting. In 25% of the surgeries a urine culture was obtained prior to surgery, though the frequency varied greatly by the service involved. Bacteruria was detected in 11% of these cultures though in most cases it was not treated. However, when they compared those patients in whom the preop bacteruria was treated with antibiotics vs. those not treated, the results were unexpected. The treated group had higher rates of surgical site infections (SSI’s) and higher rates of post-op UTI’s.
The above study had relatively small numbers and there may have been confounding factors (eg. were factors that led to the decision to treat the preoperative bacteruria also ones that led to higher rates of SSI’s?). So the study is not a definitive one but rather a hypothesis-generating one. A larger prospective study would need to be done. If it does turn out that obtaining urine cultures on admission in asymptomatic patients leads to unnecessary antibiotic treatment and that the treatment leads to an increase in post-op infections we clearly have an unintended consequence of the CMS metric of CAUTI as a never event. Thus, it may be time to retire CAUTI as the CMS metric. A more appropriate metric might simply be the percentage of urinary catheters meeting appropriateness criteria.
Continued focus on avoiding CAUTI’s is the right thing to do from a quality and patient safety perspective regardless of whether P4P or nonpayment policies are in effect or not. However, we need to make sure that using CAUTI’s as a P4P metric or flag for nonpayment is not causing more harm than good. The time may have come to retire the CAUTI metric.
Our other columns on urinary catheter-associated UTI’s:
Lee GM, Kleinman K, Soumerai SB, et al. Effect of Nonpayment for Preventable Infections in U.S. Hospitals. N Engl J Med 2012; 367: 1428-1437
U.S. Preventive Services Task Force (USPSTF). Screening for Asymptomatic Bacteriuria in Adults. July 2008
Wagenlehner FME, Naber KG, Weidner W. Asymptomatic Bacteriuria in Elderly Patients: Significance and Implications for Treatment. Drugs & Aging 2005; 22(10): 801-807
Nicolle LE, Bradley S, Colgan R, et al. Infectious Diseases Society of America Guidelines for the Diagnosis and Treatment of Asymptomatic Bacteriuria in Adults. Clinical Infectious Diseases 2005; 40: 643–54
Drekonja DM, Zarmbinski B, Johnson JR, et al. Preoperative Urine Cultures at a Veterans Affairs Medical Center. Arch Intern Med 2012; (): 1-2
Published online December 2012
Print PDF version