Abstract
This study uses electronic health record (EHR) data to evaluate differences in central line–associated bloodstream infection (CLABSI) rates by how central line days are counted: once a day at a fixed time for all patients; a sampling-based approximation; or a validated electronic count.
Central line–associated bloodstream infection (CLABSI) rates are an important quality performance indicator associated with patient morbidity, mortality, and increased costs. Hospital CLABSI rates are tracked, reported, and tied to reimbursement by the Centers for Medicare & Medicaid Services. This creates incentives for hospitals to reduce CLABSI rates with standardized best practices for the management of central lines.1,2 Between 2008 and 2017 the CLABSI standardized infection rate reported to the National Healthcare Safety Network (NHSN) fell by approximately 49% in acute care hospitals.3 The NHSN reporting guidelines specify 3 options for measuring the number of central line days: a once-a-day count at a fixed time for all patients with a central line; a sampling-based approximation to the once-a-day count; and an electronic count that may be used “after a validation of a minimum 3 consecutive months proves the data to be within 5% (+/–) of the manually collected once-a-day counts.”4 We evaluated the differences in CLABSI rates per 1000 central line days based on the choice of the time used for the once-a-day count and based on the use of electronic data collection.
Methods
This was a retrospective study at the Lucile Packard Children’s Hospital Stanford. The study was approved by the institutional review board; informed consent was waived. CLABSI data and patient central line insertion and removal times were extracted from the electronic health record for each calendar year from 2015 to 2018. These data were used to simulate 24 once-a-day counts of central line days for each hour of the day and to simulate an electronic count of any day during which a patient had a central line. The hours of the day that resulted in the lowest and highest counts were identified. The validation of the electronic count was simulated with 12 consecutive months of data from 2015. The CLABSI rates per 1000 central line days were determined for each method of counting central line days and compared by subtracting the rate based on the lowest once-a-day count from the rate based on the electronic count.
Results
Over the 4 study years, the lowest and highest number of central line days were, respectively, 30 110 and 34 320 for the once-a-day count and 32 391 and 36 030 for the electronic count. Each year, the once-a-day count was highest at 10 am and lowest at 7 pm (Figure). Over the simulated 12-month validation period, the number of line days from the electronic count was within 4.98% of the 10 am count. The number of CLABSIs per calendar year ranged from 24 to 52. Across all counts and years, the lowest and highest CLABSI rates per 1000 central line days were, respectively, 0.74 and 1.70. The largest single-year difference in CLABSI rate was –0.11 per 1000 central line days, corresponding to the difference between the electronic count and 7 pm once-a-day count in 2016 and 2017 (Table).
Figure. Central Line Once-a-Day Counts by Hour of Count.
Table. Central Line Days and CLABSI Rates per 1000 Central Line Days as Measured With Various Counts.
| Year | No. of CLABSIs | Once-a-Day Count | Electronic Count | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| At 7 pm | At 10 am | Central Line Days | CLABSIs per 1000 Central Line Days | ||||||||
| No. of Central Line Days | CLABSIs per 1000 Central Line Days | No. of Central Line Days | CLABSIs per 1000 Central Line Days | No. of Days | Difference, No. (%) | Rate | Difference, No. (%) | ||||
| From 7 pm | From 10 am | From 7 pm | From 10 am | ||||||||
| 2015 | 52 | 33 608 | 1.55 | 34 320 | 1.51 | 36 030 | 2422 (7.21) | 1710 (4.98)a | 1.44 | –0.10 (–6.72) | –0.07 (–4.75) |
| 2016 | 47 | 30 110 | 1.56 | 30 827 | 1.52 | 32 391 | 2281 (7.58) | 1564 (5.07) | 1.45 | –0.11 (–7.04) | –0.07 (–4.83) |
| 2017 | 52 | 30 614 | 1.70 | 31 423 | 1.65 | 32 826 | 2212 (7.23) | 1403 (4.46) | 1.58 | –0.11 (–6.74) | –0.07 (–4.27) |
| 2018 | 24 | 30 464 | 0.79 | 31 299 | 0.76 | 32 588 | 2124 (6.97) | 1289 (4.12) | 0.74 | –0.05 (–6.52) | –0.03 (–3.96) |
Abbreviation: CLABSI, central line–associated bloodstream infection.
Used for the simulated validation of the electronic count.
Discussion
The calculated CLABSI rate was up to 0.11 per 1000 central line days lower using electronic counts vs the lowest once-a-day counts. In 2015 and 2016, the CLABSI rates per 1000 central line days corresponding to the once-a-day 7 pm and electronic counts were, respectively, higher and lower than the Children’s Hospitals’ Solutions for Patient Safety national average of 1.50.5 Such differences could have important implications in the accuracy of reimbursement, public reporting, and hospital rankings and could create incentives for organizations to game the data.6
The primary limitations of this study are that it was performed at a single center and that it examined one particular version of electronic measurement. Other institutions may have other time-of-day differences in central line counts or other electronic methods for counting central line days.
The expansion of the adoption of electronic health records has created an opportunity for the standardization and automation of the count of central line days. Such standardization could improve the consistency and interpretability of CLABSI rates across institutions.
Section Editor: Jody W. Zylke, MD, Deputy Editor.
References
- 1.Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355(26):2725-2732. doi: 10.1056/NEJMoa061115 [DOI] [PubMed] [Google Scholar]
- 2.O’Grady NP, Alexander M, Burns LA, et al. ; Healthcare Infection Control Practices Advisory Committee . Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control. 2011;39(4)(suppl 1):S1-S34. doi: 10.1016/j.ajic.2011.01.003 [DOI] [PubMed] [Google Scholar]
- 3.Centers for Disease Control and Prevention 2017 National and state healthcare-associated infections progress report. https://www.cdc.gov/hai/data/portal/progress-report.html. Last reviewed Novemember 1, 2019. Accessed October 4, 2019.
- 4.Bloodstream infection event (central line-associated bloodstream infection and non-central line associated bloodstream infection) In: Centers for Disease Control and Prevention National Healthcare Safety Network (NHSN) Patient Safety Component Manual. Atlanta, GA: National Healthcare Safety Network; 2019. [Google Scholar]
- 5.Our results. Children’s Hospitals’ Solutions for Patient Safety website. https://www.solutionsforpatientsafety.org/our-results/. Accessed October 4, 2019.
- 6.Bastani H, Goh J, Bayati M. Evidence of upcoding in pay-for-performance programs. Manage Sci. 2019;65(3):1042-1060. doi: 10.1287/mnsc.2017.2996 [DOI] [Google Scholar]