Jennifer Meddings, Heidi Reichert, Eric Dueweke, John Rhyner
Urinary catheters often are placed unnecessarily, used without physician awareness, and associated with a very common and expensive complication—hospital-acquired catheter-associated urinary tract infection (CAUTI). In this paper we describe our experience in developing, implementing, and analyzing a retrospective, comprehensive medical record review regarding urinary catheter use and identification of urinary tract infections as CAUTIs. We share the results of urinary catheter use measures to illustrate the complexity in identifying catheter use in medical records that are being used to generate quality measures for comparing hospitals. We also discuss our experience involving resident physicians as collaborators in an opportunity for learning about patient safety research. Additionally, we share our experience related to maintaining compliance with the Accreditation Council for Graduate Medical Education (ACGME) resident duty work hour regulations.
Catheter-associated urinary tract infections (CAUTI) are among the most common healthcare-associated infections,1,2 with each having potential to cause life-threatening bacteremia and sepsis.3 With an average of one in five hospitalized patients having urinary catheters (UCs),3 UCs are a common and often avoidable hazard to patients because they are often placed unnecessarily4,5 and used without physician awareness,6,7 and may remain in use for prolonged periods,4,8 with each day increasing the risk of infection.3 Hospital-acquired CAUTI easily met criteria as a high-volume,1 expensive,3 reasonably preventable9,10 condition for which hospitals could no longer receive additional payment after the October 2008 implementation of the Hospital-Acquired Conditions (HAC) Initiative.11,12 Of note, this removal of payment for HAC began as a Medicare policy, yet rapidly expanded to many other payers, including Medicaid programs13 and Blue Cross Blue Shield14 nationwide. Hospital rates of hospital-acquired CAUTI have been publicly reported on Medicare's Hospital Compare Web site since 2011, as required by the Affordable Care Act of 2010.15
Although not paying extra for hospital-acquired CAUTIs seems like a simple concept, the HAC Initiative's implementation is complex.11,12 For a urinary tract infection (UTI) to be identified as a nonpayable, catheter-associated UTI, the claims data submitted for payment must include, in addition to the UTI diagnosis code, the ICD-9-CM [International Classification of Diseases, 9th Revision, Clinical Modification] code 996.64 for "infection and inflammation due to an indwelling urinary catheter." Both the UTI and catheter-association codes need to be identified as hospital-acquired conditions by a mandatory variable that requires all diagnoses to be identified as hospital-acquired or present-on-admission.11,12 If these codes are not assigned accurately and completely for each UTI diagnosis listed, the hospitals may receive extra payment by default.16 Similarly, inaccurate or incomplete description of CAUTI events in claims data may lead to inaccurate public reporting.17
Despite high expected rates of CAUTIs from epidemiology and surveillance studies,1,2 CAUTI rates from claims data are extremely low; many hospitals have reported zero CAUTIs despite reporting UTI rates similar to those of other hospitals. The reason for this is the lack of use of the catheter-associated code.17,18 Rare use of the catheter-associated code has been demonstrated before17 and after18 the HAC Initiative. Because hospital coders (who generate claims data) are required by Federal guidelines19 to obtain diagnoses as described in "provider" notes written by physicians, nurse practitioners, and physician assistants, we hypothesized that UTIs were not being described as catheter-associated in provider notes either because the providers were unaware of UC use, or because they did not recognize or describe UTIs as catheter-associated UTIs.
To further understand why the catheter-association code is rarely used, we performed a post-policy, retrospective, comprehensive medical record review to describe and quantify how UC use and catheter-association for UTIs are documented in different medical record types. The purpose of this paper is to share the challenges and lessons learned in the development, implementation, and analysis of this study. We share the results regarding UC use measures to highlight the complexities in detecting UC use in medical records, which are used to generate device-associated quality measures for comparing hospitals. We also share our experience involving resident physicians as collaborators in learning about patient safety research as medical record abstractors, along with our experience related to maintaining compliance with the Accreditation Council for Graduate Medical Education (ACGME) resident duty work hour regulations.20
We conducted a retrospective medical record review for a random sample of 295 adult hospitalizations, with discharges from the University of Michigan Health System (UMHS) in the first 12 months (October 1, 2008, to September 30, 2009) after the HAC Initiative's implementation. Our sample was generated by the UMHS Clinical Information and Decision Support team by first identifying all hospitalizations with UTI as a secondary diagnosis (i.e., not the primary reason for admission) in the administrative discharge data (i.e., claims) within the chosen time period (requested in four quarters). The random sample of hospitalizations for medical record review regarding catheter use was then selected using a random number generator. This project was reviewed and approved by the University of Michigan Institutional Review Board. The study included two types of data: medical records from each hospitalization and the accompanying claims data.
Comprehensive Medical Record for Hospitalization
This retrospective medical record review was performed from May 2009 to May 2011, a period in which hospitalization records were all accessible electronically but used different computer systems and methods, depending on the type of documentation. Provider notes were entered by dictation or typing into an electronic medical record (EMR) system called CareWeb; nurse notes were usually scanned into CareWeb from paper bedside flowsheets; emergency department (ED) provider orders were scanned-in documents; inpatient orders were recorded in the computerized provider order entry (CPOE) system called CareLink (implemented April 2008); and test results were searchable in CareWeb. Figure 1 illustrates the type and sequence of document review performed. To facilitate a systematic review of these document types, an electronic survey-generating tool was adapted to serve as a guide to remind reviewers how to locate each type of documentation, to require responses to specific questions (Appendix, Table A-1) about the documentation regarding catheter use (all UC types: Foley, intermittent straight catheter (ISC), nephrostomy, suprapubic, and external, as defined in Appendix, Table A-2) and association with UTIs. Skip logic allowed the abstractor to skip questions not relevant to a particular record (e.g., ED-related questions if there was no ED course). Abstractors also answered some questions using any available hospitalization records regarding catheter use and UTI development. Though beyond the scope of this paper, data were collected regarding documented UTI symptoms and laboratory data in a manner to permit categorization of UTIs by various diagnostic criteria.9,10
We chose resident physicians as abstractors because we thought this project would be an excellent opportunity for residents to learn about patient safety research, and because resident physicians had the medical and practical expertise and experience needed to review all required documents.
The principal investigator (JM) met with the internal medicine residency program director to request involvement of residents; it was decided the abstractor positions could serve as internal "moonlighting" opportunities for two second-year residents in good academic standing whose rotation schedules in the project's timeline were expected to have a few hours each week available within the ACGME duty hour restrictions.20 "Moonlighting" describes work opportunities where licensed residents can earn income beyond their resident salary. Residents were limited to working 80 hours per week (including internal moonlighting20) and were provided at least one 24-hour day in 7 days that was free from all clinical, educational, and administrative duties. Residents were required to log all work hours (including moonlighting) into an electronic system called MedHub, and a detailed protocol was followed regarding submission of work hours for review to the residency program and the graduate medical education (GME) office. The physician-abstractors underwent 4 hours of data collection training; they provided feedback that was incorporated into the final abstraction tool. They also completed the University of Michigan's Program for Education and Evaluation in Responsible Research and Scholarship. Residents met with JM periodically during and after completing the medical record abstraction to share their experience and challenges. After completion of the abstractions by resident-physicians using the electronic tool, a third physician (JM) re-abstracted provider notes to assess details of the language used to describe UC use and catheter association for UTIs; this was necessary after clarification from our hospital coders regarding the type of language required in provider notes to prompt application of the catheter-association code to UTI diagnoses in claims data.
Figure 1. Comprehensive medical record review flow diagram
In addition to being an organized, complete, and reproducible review of the medical record, the abstraction process was designed to be flexible to accommodate the resident physicians' schedules. The physician-abstractors accessed the medical records electronically (using the same data security standards that they used daily in accessing the records for patient care) and were guided by the password-protected, encrypted electronic tool to systematically review and answer questions from the medical record regarding UC use and UTI development. Abstractors were provided a list of medical record numbers and discharge dates (in password-protected, encrypted files), each linked to a new chart review number (based on abstractor and review number), to identify the hospitalization record in the abstracted data file. To optimize use of abstractor time and data security, the abstraction was designed specifically to collect only categories of data (such as UC use and UTI development), without any patient or hospitalization identifiers.
Administrative Discharge Abstract (Claims Data)
After completion of all medical record abstractions, claims data were requested. Claims data included all diagnosis and procedure codes applied by the hospital coders to request payment for the hospitalization after the patient was discharged, as well as routine patient demographics. Claims data also included the new mandatory variable required by the HAC Initiative,12 which required all diagnoses to be identified as hospital-acquired or present-on-admission. Comorbidity variables21 were generated from the claims data using comorbidity software (version 3.4) from the Agency for Healthcare Research and Quality (AHRQ). By inclusion criteria, all hospitalizations were for adults not admitted to obstetrics, whose claims data included at least one UTI code as a secondary diagnosis from the 10 diagnosis codes chosen for nonpayment in the HAC Initiative12 (i.e., 112.2, 590.1, 590.11, 590.2, 590.3, 590.80, 590.81, 595.0, 597.0, 599.0). There were two potential indicators of UC use in claims data: (1) the 996.64 catheter-association code to identify UTIs as CAUTIs and (2) inclusion of an ICD-9-CM procedure code for insertion (57.94) or replacement (57.95) of an indwelling urinary catheter. Of particular importance, UC procedure codes are applied only when UCs are placed by physicians; however, nurses place the majority of UCs.
Descriptive summary statistics are reported. Data management tasks for the abstracted data in survey format were performed using SPSS, version 18 (SPSS Inc., Chicago, IL). The merged dataset of abstracted medical record data with claims data was analyzed using Stata/MP, version 12.1 (StataCorp, College Station, TX). Thirty records were reviewed by both physician-abstractors to assess inter-rater reliability using the kappa statistic. A random number generator was used to select one record from each of the 30 duplicate abstracted records to be included in the analytic sample.
In all, 295 records were requested for this sample, but one hospitalization had been converted to an outpatient visit after our original sample was created and therefore did not have accompanying inpatient claims data. Our random sample of 294 hospitalizations (for 289 unique patients, including 5 patients with 2 hospitalizations in the sample) that included a secondary diagnosis of UTI had the following characteristics: 193 (65.7 percent) were for women; mean age 63 years (range: 22–98), median length of stay 7.5 days (IQR 1–16, range 1–84), including 181 (61.6 percent) with Medicare; and 21 (7.1 percent) deaths while admitted. Common comorbidities included renal failure (23.5 percent), chronic lung disease (16.7 percent), paraplegia (11.2 percent), other neurologic diseases (13.3 percent), diabetes (10.2 percent), and heart failure (9.5 percent). A total of 163 (55.4 percent) patients had ED evaluations before admission.
Measures of Urinary Catheter Use
Urinary Catheter Use Documented in Provider Notes
Table 1 summarizes the quantitative measures of UC use from provider notes reviewed for each time period in the hospitalization, including ED course, admission day, post-admission days (defined as day after admission until day before discharge), and discharge day. Examples of evidence of UC use in provider notes included (1) patient interview information, such as "Foley since surgery 4 days ago" or "complains Foley hurts"; (2) exam findings, such as "bilateral nephrostomy tubes are present"; (3) provider's review of tests, such as "urinalysis from Foley has bacteria"; and (4) provider's assessment and plan, such as "UTI, due to Foley" or "ISC every 6 hours if no void." References to remote UC use (e.g., "patient required Foley catheter temporarily after prostate surgery in 2005") were not included in this measure. Overall, provider notes identified 184 patients (63 percent) with at least one provider note describing UC use (as current use or ordered, or by mention of patient having a catheter-associated UTI).
Although not quantified in this study, the detailed abstraction of language used in provider notes to describe catheters and catheter association for UTIs suggested some patterns of provider documentation regarding catheters. It was not unusual for the first mention of a UC in provider notes to be in a consultant note well into the hospitalization. Consultation notes commonly describing UCs were from physical medicine and rehabilitation, the wound/ostomy care team, urology, and geriatrics; catheter placement was often requested in nephrology consultation notes for obtaining 24-hour urine volumes and test results. Unexpectedly, infectious disease consultations did not frequently comment regarding ongoing UC use as part of the assessment if it was not related to the reason for consultation. Primary managing teams, whose notes often mentioned UC use, included hospitalists and intensive care teams, usually in a dedicated section of the note for devices; highly structured notes (that appeared to have been typed using a template) seemed more likely to mention UC use. Some provider notes with dedicated device sections included vascular lines (such as central venous catheters) without noting UCs in use by nursing flowsheets. In this sample, providers from the anesthesia teams managing surgical intensive care unit (ICU) patients seemed the most reliable documenters of UC presence or absence. If UC use was mentioned once in a provider note, it commonly was mentioned on a recurring basis. Sometimes this appeared to be by exact copying of text from one day's note to another; at other times it appeared to be due to increased awareness of UC use after it was first noted.
Table 1. Urinary catheter (UC) use as documented throughout hospitalization course
|Type of documentation||Time Period of Hospitalization|
|Provider Notes describing UC use
Overall, provider notes identified 184 (63%) hospitalizations with UC use.
|ED providers described UC use for 31 of the 163 hospitalizations with an ED course.||Admitting providers described UC use for 91 (31%) hospitalizations.||Post-admission provider notes described UC use for 164 hospitalizations.||Provider notes on discharge day described UC use during 62 (21%) hospitalizations.
Provider notes describe expected post-discharge UC use for 28 (10%) hospitalizations.
|106 (36%) of patients had UC use mentioned in provider documentation for ED course or admission day.|
|Nursing documentation of UC use
Overall, bedside nursing flowsheets identified 212 (72%) hospitalizations with UC use.
|Not available to review||Nurses noted UC use for 143 (49%) hospitalizations on bedside flowsheet on admission day.||Nurses noted UC use for 204 (69%) hospitalizations by bedside flowsheet on post-admit days.||Nurses noted UC use for 91 (31%) hospitalizations by bedside flowsheet on discharge day, with 28 (9.5%) UCs removed on the discharge day.|
|By nursing inpatient flowsheets (available for 289 [98%] hospitalizations), 212 hospitalizations had ≥1 UC from day of admission until discharge including 180 (61%) Foley UCs, 63 (21%) ISCs, 5 (2%) external UCs, 9 (3%) nephrostomy UCs, 5 (2%) suprapubic UCs.|
|Provider orders regarding UCs
Overall, provider orders regarding UC identified 222 (76%) hospitalizations with UC orders
|ED providers ordered UCs for 24 hospitalizations, including 18 Foley UCs and 6 ISCs.||Using the inpatient electronic CareLink order system (all orders from admission to discharge):
|Claims data indicators of UC use
Overall, claims data codes identified 25 (8%) hospitalizations with UC use.
Abbreviations: ED=emergency department; UC=urinary catheter (any type unless specified); ISC=intermittent straight catheter UC ICD-9-CM=International Classification of Diseases, 9th Revision, Clinical Modification.
Notes: Post-admission course is defined as time period after admission day but before discharge day.
Claims data contain diagnosis and procedure codes describing the entire episode of care (ED course through discharge)
Urinary Catheter Use Documented in Nurse Notes
Evidence of UC use in nurse notes primarily came from daily flowsheets monitoring fluid intake and output (often with abbreviations such as a circled "F=300," meaning 300 cc urine output in Foley catheter bag) and periodic assessments of continence devices. Overall, nurse notes identified 212 hospitalizations with at least 1 day's record indicating UC use, including 143 (49 percent) on admission day and 204 (69 percent) on post-admission days. On discharge day, nurse notes indicate that 91 (31 percent) patients had UCs used on the day of discharge, including 28 percent with the catheter removed on the day of discharge.
Urinary Catheter Use Documented in Provider Orders
ED providers ordered UCs impacting 24 hospitalizations, including 18 Foley catheters and 6 ISC orders. By the inpatient CPOE orders, 211 hospitalizations had orders to place (including "as needed" ISC orders), maintain, or discontinue UCs, including 176 hospitalizations with Foley catheters and 103 with ISC orders. Overall (with a kappa of 0.9), 222 (76 percent) hospitalizations were noted to have at least one UC order by either an ED or inpatient provider.
Urinary Catheter Use Documented in Administrative Discharge Abstract (Claims Data)
Claims data contained the ICD-9-CM catheter-association code 996.64 for 20 (6.8 percent) hospitalizations, including 11 with diagnosis codes describing hospital-acquired catheter-associated UTIs. ICD-9-CM procedure codes for inserting or replacing indwelling urinary catheters were listed for five hospitalizations. Overall, claims data identified 25 (8 percent) hospitalizations with UC use identified by either the catheter-association or catheter placement codes.
Frequency of Provider Identification of UTIs as CAUTIs
Given the rare use of the catheter-association code 996.64 by hospital coders to describe UTIs as CAUTIs in claims data, and the requirement for hospital coders to obtain diagnoses such as UTI for claims from provider notes, we evaluated the language used by providers in their documentation to describe UTIs as catheter-associated or not. From conversations with our hospital coders, physicians needed to clearly describe the UTI as being catheter-associated; it was not sufficient to describe a UTI and catheter use individually, such as "UTI, plan: remove Foley." Coders look for specific provider language such as "UTI due to catheter," "UTI due to Foley," or "catheter-associated UTI." In our sample, providers used language indicating that UTIs were catheter-associated for 22 (7.5 percent) hospitalizations, as categorized in Table 2. Of note, for 5 of these 22 hospitalizations, the discharge summaries had addenda that included a clarification regarding a diagnosis of catheter-associated UTI, including 4 hospitalizations for which the addenda accounted for the only mention of a catheter-associated UTI in the provider notes.
Table 2. Language in provider notes (for 22 hospitalized patients) to describe catheter-associated UTIs
|Phrases used by providers to describe CAUTIs||Number* of patients for whom phrase was used|
|The phrase "catheter-associated UTI or "catheter-associated urinary tract infection" or "UTI due to catheter" was used||4|
|Abbreviation "CAUTI" or "CA-UTI" or "CA-urinary tract infection" was used||0|
|Provider documented the word "Foley" in association with a UTI diagnosis, such as "UTI due to Foley"||13|
|Provider specifically mentioned another type of urinary catheter in association with UTI (such as "UTI due to nephrostomy" or "suprapubic catheter UTI"||6|
*Numbers sum to 23 because for one patient with provider-described CAUTI, two different types of phrases were used to describe CAUTI in the medical record.
Note: CAUTI=catheter-associated urinary tract infection; UTI=urinary tract infection.
Variations in Identification of UC Use by Type of Documentation
As Figure 2 illustrates, the four different documentation types identified slightly different patient groups as having UC use. Overall, 235 (80 percent) hospitalizations in the sample were identified as having UC use by at least one of the four types of documentation, with 166 hospitalizations identified by provider notes, nurse notes, and provider orders and 22 hospitalizations identified by all four documentation types including claims data.
Our single-site study found that the majority (80 percent) of hospitalized patients discharged with a secondary diagnosis of UTI had evidence of UC use in the medical record. Completing this comprehensive medical record review was resource intensive and time consuming (requiring an average of 1 hour per hospitalization) and required expertise in accessing, navigating, and reviewing clinical documentation fragmented across multiple EMRs. Reviewing provider notes and nurse bedside flowsheets was the most resource-intensive activity. Reviewing UC use by provider orders in the CPOE system was simpler and identified the most hospitalizations (n=222, 76 percent) with UC use; however, some provider orders included "as needed" criteria and so may not reflect actual catheter use. Given that our CPOE system was relatively new and that UCs had previously been placed without physician orders6 or awareness,7 we were pleasantly surprised that 207 (94 percent) of all 220 hospitalizations (Figure 2) with evidence for UCs in provider or nurse notes did have a provider order involving UCs. Nurse notes identified 212 (72 percent) patients with UCs, which reflected actual catheter use because nurses documented their tasks performed using the UC (such as amount of urine noted in a Foley catheter bag, or urine obtained using an ISC). Electronic systems for nurse documentation (such as Centricity Clinical Information View [GE Medical Systems, 2002], whose use has since expanded at UMHS) can be queried in less time; still, any electronic entry by busy clinicians should validated due to possible inaccuracies from default responses and copy/paste issues. In summary, except for reviewing provider orders, reviewing the multiple inpatient medical record documentation systems for catheter use was akin to hunting for a needle in a haystack due to the sheer volume of records requiring manual review and inconsistent documentation of UC use in provider notes.
Figure 2. Venn diagram of urinary catheter use indicated by four types of documentation: provider notes, inpatient nurse notes, provider orders, and claims data
This study has some important limitations, primarily the fact that only one academic medical center was examined, using a modest number of hospitalizations. However, the rare use of the catheter-association code in our sample is similar to low rates seen in statewide and nationwide claims data.17,18 This study was not designed to determine why provider notes rarely describe UTIs as catheter-associated. Informal queries with some of our physicians (residents and hospitalists) suggest two reasons: (1) providers are unaware how strictly their documentation is reviewed by hospital coders selecting diagnosis codes to describe UTIs, and (2) in comparison to a patient's other medical problems requiring treatment, catheter use or UTI may be a lower priority for the time dedicated to documentation. Because UMHS providers have no personal disincentive for describing CAUTIs in their notes, we do not believe providers intentionally avoided describing UC use or catheter association for UTIs.
The resident physicians' experience and insights were invaluable to this project, particularly given the complexities of the EMRs. Both abstractors enthusiastically expressed that it was an excellent opportunity to learn about patient safety research, including device use, device-associated complications, and implications of physician documentation. One abstractor stated, "In short, this project was a far more effective patient safety curriculum than many other residents had available to them." The residents also stated that their experience was advantageous in their application to competitive fellowships and described how the project impacted their career interests:
- "I don't think it can be stressed enough how useful this experience was to me … the project helped to fine tune my career interests. I plan to make quality and process improvement a large part of my career."
- "My involvement in this project … was an important stepping stone in both clarifying and strengthening my interest in quality improvement and advancing my career in that direction."
These physicians also appreciated the "moonlighting" opportunity to help supplement their income while also contributing to their education:
- "We had the benefit of being engaged and inspired by our own faculty while participating in meaningful research, while simultaneously supplementing our income at a time when dollars matter."
- "For many … time spent as a house officer is period of relative financial desperation. The pressure to augment income is strong. To have the opportunity to do it while simultaneously performing meaningful work and learning about something relevant is a gift."
However, resident involvement in this project was not without challenges. Although we anticipated that hiring residents would be somewhat more complicated than other abstractors, given resident schedules and ACGME duty hour regulations,20 we did not anticipate the tremendous complexity of the review process required before the resident could be paid. Compared to other moonlighting options, this project was more complex to administer because the work was performed on a very flexible schedule, with the amount of time spent in one "shift" varying from 15 minutes to several hours. The process for submitting timesheets to release payment required several modifications to address unexpected challenges. The finalized process was as follows:
- Resident entered the time worked into an electronic system called MedHub.
- Resident submitted "moonlighting" paper timesheet with the days/hours worked to the research team "moonlighting supervisor" (JM) to review, sign, and enter the research account code from which payment would come.
- The supervisor sent the signed timesheet to the residency program office for review.
- Residency program office reviewed, queried if needed, approved, signed the timesheet, and returned it to the supervisor.
- The supervisor submitted the completed timesheet to the GME office.
- The GME office also reviewed the duty hours for compliance and, if there were no discrepancies, submitted the approved timesheet to payroll to generate payment from the research account.
Several steps involved time limitations as to when the process had to be completed to progress to the next stage. Other challenges included delays in timesheet approval due to unexpected reports of "discrepancies" (even though a work hour violation had not occurred) such as:
- Work recorded in the paper timesheet in minutes did not exactly match the duty hours recorded in MedHub (which rounded to 30-minute blocks).
- MedHub would not allow recording of moonlighting hours on residency vacation days, generating discrepancies between the paper timesheets and electronically recorded hours.
- Need for duty hour assessors (particularly with staff changes) to recognize that moonlighting hours could be contiguous with usual resident work hours (e.g., performing medical record reviews from 5 p.m. to 7 p.m. after working an 8 a.m. to 5 p.m. residency shift).
Without a doubt, the most important challenge that residents faced was stress from responding to multiple queries regarding these work hour "discrepancies," with potential for their participation ending or pay denials while the research team, residency office, and GME office addressed each new issue that arose. Fortunately, despite these unexpected and time-consuming challenges, the medical record review process was completed, with the resident-physicians paid in full before completing their residency.
In summary, this single-site study involving a comprehensive medical record review of patients discharged with a UTI diagnosis found that although UC use remains very common in this patient population and was documented routinely by nurses and in the electronic orders, it was more difficult to detect in provider notes, which are the primary data source that hospital coders review to generate the claims data. Although provider notes often commented on the use of UCs, providers rarely described UTIs specifically as catheter-associated diagnoses using the text that hospital coders require to describe UTIs as CAUTIs in claims data. This likely explains why so few UTIs are identified as CAUTIs in claims data, despite the fact that CAUTIs remain a common hospital-acquired infection. This significant problem with identification of catheter-association for UTIs in claims data supports the recent decision, in June 2013,22 to use CAUTI rates from surveillance data collection reported to the National Healthcare Safety Network (and also to Hospital Compare) instead of CAUTI rates from claims data as the measure for comparing hospital performance by complication rates and assigning financial penalties, beginning in October 2014. However, hospital rates of CAUTI from claims data are still being publicly reported on Hospital Compare(as of August 2013), despite the results of this study in combination with others,17,18 indicating that claims data are not a reliable data source of either UC use23 or identification of CAUTI events.17,18
We conclude by sharing some principles and lessons learned from conducting this project.
- Urinary catheter use remains a common risk factor for many hospitalized patients and was most frequently documented in this medical record review by electronic orders (which required little time or resources to review) and nurse flowsheets (whose review was time intensive in this study because of the use of scanned bedside flowsheets, but could be simplified using electronic nurse data entry systems).
- Physicians rarely describe UTIs with language specific enough to be interpreted by hospital coders as being catheter-associated UTIs, despite often acknowledging UC use in notes.
- Research tasks (such as this comprehensive medical record review) can be designed to be flexible, educational, inspiring, and rewarding opportunities for resident physicians. However, involving resident physicians as moonlighting collaborators has some unique challenges that may be more difficult to address as ACGME duty hour requirements evolve over time.
Funding for this project was generously provided in part by the Blue Cross Blue Shield of Michigan Foundation (1452.11, PI: Meddings) and the Agency for Healthcare Research and Quality, U.S. Department of Health and Human Services (K08 HS19767-02, PI: Meddings). The Blue Cross Blue Shield of Michigan Foundation grant provided the resources for the project's design and data collection by the resident physician-abstractors; the AHRQ K08 further supported Dr. Meddings while she performed the detailed re-abstraction of provider notes regarding language used by providers related to catheter use and UTIs. The findings and conclusions in this document are those of the authors, who are responsible for its content, and do not necessarily represent the views of AHRQ. No statement in this report should be construed as an official position of AHRQ or of the U.S. Department of Health and Human Services.
We wish to express our appreciation to the University of Michigan Internal Medicine Residency Program for allowing our resident physician team members (JR, ED) to participate as physician-abstractors for this project and to the Graduate Medical Education office for performing the task of resident work hour monitoring required for compliance with the ACGME requirements. We acknowledge the assistance of Vinita Bahl, DMD, MPP, in obtaining our medical record sample and administrative data. We appreciate the insight provided by Gwendolyn Blackford, BS, RHIA, regarding processes used and regulations followed by hospital coders to assign diagnosis codes in claims data.
University of Michigan Medical School, University of Michigan Health System, Department of Internal Medicine, Division of General Medicine, Ann Arbor, MI (JM, HR). University of Pittsburgh Medical School and University of Pittsburgh Medical Center, Department of Medicine, Division of Cardiology, Pittsburgh, PA (ED). Northwestern University Feinberg School of Medicine, Northwestern Memorial Hospital, Department of Medicine, Division of Cardiology, Chicago, IL (JR).
Address correspondence to: Jennifer Meddings, M.D., M.Sc.; Department of Internal Medicine, Division of General Medicine, 2800 Plymouth Road, NCRC Building 16, Room 430W, Ann Arbor, MI, 48109; Email: firstname.lastname@example.org.
1. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J Med 2002;Jul 8;113 Suppl 1A:5S-13S. PMID: 12113866.
2. Dudeck MA, Horan TC, Peterson KD, et al. National Healthcare Safety Network (NHSN) Report, data summary for 2010, device-associated module. Am J Infect Control 2011 Dec;39(10):798-816. PubMed PMID: 22133532.
3. Saint S. Clinical and economic consequences of nosocomial catheter-related bacteriuria. Am J Infect Control 2000 Feb;28(1):68-75. PMID: 10679141.
4. Jain P, Parada JP, David A, et al. Overuse of the indwelling urinary tract catheter in hospitalized medical patients. Arch Intern Med 1995 Jul 10;155(13):1425-9. PMID: 7794092.
5. Fakih MG, Watson SR, Greene MT, et al. Reducing inappropriate urinary catheter use: a statewide effort. Arch Intern Med 2012 Feb 13;172(3):255-60. PubMed PMID: 22231611.
6. Saint S, Wiese J, Amory JK, et al. Are physicians aware of which of their patients have indwelling urinary catheters? Am J Med 2000 Oct 15;109(6):476-80. PMID: 11042237.
7. Fakih MG, Shemes SP, Pena ME, et al. Urinary catheters in the emergency department: very elderly women are at high risk for unnecessary utilization. Am J Infect Control 2010 Nov;38(9):683-8. PMID: 21034978.
8. Fakih MG, Dueweke C, Meisner S, et al. Effect of nurse-led multidisciplinary rounds on reducing the unnecessary use of urinary catheterization in hospitalized patients. Infect Control Hosp Epidemiol 2008 Sep;29(9):815-9. PMID: 18700831.
9. Hooton TM, Bradley SF, Cardenas DD, et al. Diagnosis, prevention, and treatment of catheter-associated urinary tract infection in adults: 2009 International Clinical Practice Guidelines from the Infectious Diseases Society of America. Clin Infect Dis 2010 Mar 1;50:625-63. PMID: 20175247.
10. Gould CV, Umscheid CA, Agarwal RK, et al. Guideline for prevention of catheter-associated urinary tract infections 2009. Infect Control Hosp Epidemiol 2010 Apr;31(4):319-26. PMID: 20156062.
11. Centers for Medicare & Medicaid Services (CMS), HHS. Medicare program; changes to the hospital inpatient prospective payment systems and fiscal year 2008 rates; hospital-acquired conditions, including infections. Fed Regist 2007 Aug 22;72(162):47200-18. PMID: 17847578.
12. Centers for Medicare & Medicaid Services (CMS), HHS. Medicare program; changes to the hospital inpatient prospective payment systems and fiscal year 2009 rates. Fed Regist 2008 Aug 19;73(161):48473-91. PMID: 18956499.
13. Centers for Medicare & Medicaid Services (CMS), HHS. Medicaid program; payment adjustment for provider-preventable conditions including health care-acquired conditions. Fed Regist 2011 June 6;76(108):32816-38.
14. Blue Cross Blue Shield Association. Blue Cross and Blue Shield announces system-wide payment policy for "never events." Press release. Washington, DC; February 26, 2010.
15. Affordable Care Act, Section 3008: Payment adjustment for conditions acquired in hospitals. Pub L No. 111–148, 124 Stat 122. Available at: www.gpo.gov/fdsys/pkg/PLAW-111publ148/pdf/PLAW-111publ148.pdf (PDF File, 141 KB). Accessed July 10, 2012.
16. Saint S, Meddings JA, Calfee D, et al. Catheter-associated urinary tract infection and the Medicare rule changes. Ann Intern Med 2009 Jun 16;150(12):877-84. PMID: 19528567.
17. Meddings J, Saint S, McMahon LF Jr. Hospital-acquired catheter-associated urinary tract infection: documentation and coding issues may reduce financial impact of Medicare's new payment policy. Infect Control Hosp Epidemiol 2010 Jun;31(6):627-33. PMID: 20426577.
18. Meddings JA, Reichert H, Rogers MA, et al. Effect of nonpayment for hospital-acquired, catheter-associated urinary tract infection: a statewide analysis. Ann Intern Med 2012 Sep 4;157(5):305-12. PMID: 22944872.
19. Centers for Medicare & Medicaid Services (CMS), National Center for Health Statistics (NCHS). ICD-9-CM Official Guidelines for Coding and Reporting. Effective 1 October 2008 (Fiscal Year 2009). Available at: www.ama-assn.org/resources/doc/cpt/icd9cm_coding_guidelines_08_09_full.pdf. Accessed January 29, 2013.
20. Accreditation Council for Graduate Medical Education (ACGME). Frequently asked questions—ACGME common duty hour requirements, effective July 1, 2011.
21. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care 1998 Jan;36(1):8-27. PMID: 9431328.
22. Centers for Medicare & Medicaid Services (CMS), HHS. Medicare program; hospital inpatient prospective payment systems for acute care hospitals and the long term care hospital prospective payment system and proposed fiscal year 2014 rates. Fed Regist 2013 May 10;78(91):27622-35.
23. Zhan C, Elixhauser A, Richards CL Jr, et al. Identification of hospital-acquired catheter-associated urinary tract infections from Medicare claims: sensitivity and positive predictive value. Med Care 2009 Mar;47(3):364-9. PMID: 19194330.
Table A-1. Examples of questions in the electronic tool to guide medical record review
|Section: Emergency Department Provider Notes in CareWeb
Please go to CareWeb "Documents" → "CareWeb Docs" for this hospitalization, and answer the following questions using notes labeled "ED NOTE" and "CONSULT-ED."
|Is there emergency provider documentation (ED NOTE or CONSULT-ED) for this hospitalization?
___ No (if No, tools skips to questions regarding Inpatient Admission notes)
|According to documentation available in these ED provider notes, please indicate what information supported that this patient has a urinary tract infection (UTI) at the time of admission to the hospital from the ED?
Please check ALL that apply.
___ Vital signs T>100.4 F or >38 C, with no other recognized cause than UTI*
___ Patient/caregiver report of fever or chills/rigors†
___ Urinary complaint of dysuria, frequency, or urgency*
___ Suprapubic tenderness or costovertebral angle pain or tenderness*†
___ Altered mental status, malaise or lethargy†
___ New or worsening incontinence**
___ For spinal cord injury patients: increased spasticity or autonomic dysreflexia†
___ Urine described as malodorous or discolored**
___ Abnormal urinalysis (UA) known to ED team: no details given
___ Abnormal UA: positive leukocyte esterase or nitrite*
___ Abnormal UA: pyuria (≥10 WBC/mm^3 or ≥3 WBC/HPF of urine)*
___ Abnormal UA: microorganisms seen on Gram stain of urine*
___ Positive urine culture known to ED team: no details given
___ Positive urine culture: ≥10^5 CFU/mL with ≤2 species of bacteria*
___ Positive urine culture: ≥10^3 and <10^5 CFU/mL with ≤2 species of bacteria*
___ Positive urine culture: <1000 CFU/mL yeast
___ Positive urine culture: ≥1000 CFU/mL yeast*
___ Positive blood culture for gram-negative infection
___ Positive blood culture, other organism
___ ED provider listed UTI as a problem or suspected diagnosis for this admission††
___ ED provider initiated/continued empiric antimicrobial treatment for suspected or confirmed UTI††
___ NONE, no information in ED provider notes supports UTI diagnosis was suspected or possible
___ Other information supporting UTI diagnosis, not listed above (such as radiology test supporting infection, or culture from non-urine fluid or tissue*):
|What kind of urinary catheter was being used at the time of presentation to the ED?
___ NONE, no evidence a urinary catheter was being used at the time of presentation to ED
___ Foley: indwelling trans-urethral catheter
___ Suprapubic: indwelling bladder catheter, not trans-urethral
___ Nephrostomy tube: indwelling, into kidney or ureter
___ External catheter: "condom" catheter for male patient
___ Intermittent straight catheter (ISC) use
___ Urinary catheter in use, but cannot determine type from ED note
___ Other urinary catheter in use: ___________________________
|Section: CareWeb Lab results
What were the URINALYSIS results when collected as part of the ED or inpatient admission work-up?Urinalysis test codes are usually: UA, UMIC, UMAC. When using the search function, please search codes individually. Please check ALL that apply.
___ Urinalysis not collected in ED or as part of admission work-up.
___ Urinalysis results were NOT suggestive of UTI (no leukocyte esterase, no nitrite, no WBCs, no WBC casts, no microorganisms)
___ Positive leukocyte esterase
___ Positive nitrite
___ Pyuria (≥10 WBC/mm^3 or ≥3 WBC/HPF of urine)
___ Microorganisms seen on Gram stain or urine
___ Other (please specify): _____________________
|What were the urine culture results when collected as part of ED or inpatient admission work-up? Urine culture test code is usually URCC. Please include results for both bacteria and yeast growth from urine culture. Please check ALL that apply.
___ No urine culture was collected as part of ED or inpatient admission work-up
___ Urine culture results had NO growth
___ Positive urine culture of ≥10^5 CHF/mL with no more than 2 species of bacteria in sample tested
___ Positive urine culture of ≥10^3 and <10^5 CFU/mL with no more than 2 species of bacteria in sample tested
___ Positive urine culture: <1000 CFU/mL yeast
___ Positive urine culture: ≥1000 CFU/mL yeast
___ Other (please specify):_________________________
|Section: Review of CareWeb "Imaged Documents" for this Hospitalization
Refer to "UMHHC Chart—Inpatient Chart—Orders (Patient Care/Diagnostic/PCA/Epidural/PN/ATF": this is where you find physician orders for ED course.
|Yes||No||Not applicable, no ED course for patient|
|Were orders given for a urinalysis (UA/"dip")?||___||___||___|
|Were orders given for Foley catheter insertion?||___||___||___|
|Were orders given for condom catheter placement?||___||___||___|
|Were orders given for straight catheter use?||___||___||___|
|Were orders given for suprapubic catheter placement or use?||___||___||___|
* Part of 2009 Centers for Disease Control (CDC)10 criteria for identifying symptomatic urinary tract infections, when interpreted in context of patient age and whether urinary catheter in place currently or in past 48 hours.
† Part of 2009 Infectious Diseases Society of America (IDSA)9 criteria for identifying symptomatic urinary tract infections when interpreted in context of catheter type, specimen collection type, and whether urinary catheter is in place currently or in past 48 hours.
** Other common criteria that physicians use to identify urinary tract infections.
†† Criteria hospital coders can use as evidence of UTI documentation in provider notes †.
Table A-2. Urinary catheter type definitions
|Catheter Type||Urinary catheter description|
|Foley catheter||The Foley catheter is an indwelling trans-urethral urinary catheter that is a flexible plastic tube passed through the urethra into the bladder to drain urine from the bladder to be collected in a urine storage bag. This catheter can be placed either for temporary use (hours–days) or prolonged use (weeks–months). These catheters are most often placed by nurses but occasionally are placed by physicians (such as urologists) when placement is challenging for anatomic reasons. Most Foley catheters are used for short periods of time while a patient is hospitalized or having an outpatient procedure, but some patients with chronic medical issues have Foley catheters in place at home, outside of the hospital. "Foley" catheters are named for Dr. Frederic Foley, the urologist who designed them.|
|Intermittent straight catheters (ISC)||Intermittent straight catheters (ISC) are non-indwelling catheters that are stiffer plastic tubes passed through the urethra into the bladder to drain urine from the bladder. Unlike the Foley catheter, the ISC does not remain in the bladder; instead, it is used to periodically empty the bladder and is in place for only a few minutes (removed after flow of urine ceases), using a technique of "in-and-out" catheterization of the bladder. Patients may require use of an ISC only on an as-needed basis (such as in a full bladder requiring drainage once while perioperative medications make spontaneous bladder emptying difficult). Other patients require multiple catheterizations using an ISC per day on a scheduled basis (such as in patients with chronically enlarged prostates). ISC catheters are most often placed by nurses but can also be placed by patients performing self-catheterization.|
|Nephrostomy catheter||A nephrostomy catheter is a type of indwelling catheter that is a flexible tube that is placed through the skin into the kidney to drain urine that cannot be urinated through the bladder due to a blockage in the ureter tubes, which drain urine from the kidney to the bladder. Nephrostomy tubes are placed by surgical physicians or interventional radiologists, most often in operating rooms. Nephrostomies can be placed temporarily (for example, to drain urine until a kidney stone is removed) or can be placed for long-term use, including in patients being discharged home.|
|Suprapubic catheters||A suprapubic catheter is a type of indwelling catheter that is a flexible tube surgically placed through the skin into the bladder to drain bladder urine that for anatomic or practical reasons cannot be managed by another type of catheter. This catheter can be the same as a Foley catheter, but it is named suprapubic to designate its location. Suprapubic catheters are initially placed and changed as needed by physicians, but after the patient has had a suprapubic catheter in place for a while, it can be changed by a nurse or the patient.|
|External "condom" catheters||An external catheter is a non-indwelling urine collection device that consists of a flexible tube and urine storage bag (similar to a Foley catheter) attached to the patient by a condom that is fit over the penis. Although similar devices have been tested for female patients, currently external catheters are used primarily in male patients. This type of catheter is used when there is a need for urine collection, but the patient has no difficulty releasing urine from the bladder. External catheters are also commonly known as "Texas catheters" because they are often manufactured in Texas.|
|Other urine collection devices that are NOT catheters include:
Urostomy: a surgically created opening in the abdominal wall to allow drainage of urine from the kidneys and ureters into the surgically created new pathway (called an ileal conduit, created using a small segment of bowel) to reach the opening on the skin. Urine is collected from the opening into a "pouch," which is a type of plastic bag attached to the skin by an adhesive, changed periodically to empty the urine.