Evaluation of a Resident-Led Project to Decrease Phlebotomy Rates in the Hospital: Think Twice, Stick Once

Excessive inpatient laboratory testing leads to unnecessary health care cost and exposes patients to uncomfortable blood draws, false-positive results, and hospital-acquired anemia.^1,2

Many strategies have sought to decrease laboratory testing in the hospital,³ but most do not focus on decreasing needlesticks for patients.

In July 2014, we launched the resident-led Think Twice, Stick Once program on the Internal Medicine teaching service at the University of California, San Francisco (UCSF) Medical Center to reduce the number of phlebotomies per patient per day by at least 5%.

We designed the program to follow the previously described Culture, Oversight, Systems Change, Training (COST) framework for value-improvement projects.⁴ The campaign featured resident-physician and hospital leaders and raised awareness of the program through posters, buttons, and pens with the slogan, “Think Twice, Stick Once.” Team-specific phlebotomy data were reported twice monthly to members of the medicine teams. The project was supported by the UCSF’s Housestaff Quality Improvement Incentive Program,⁵ which provides each resident with an annual $400 reward for meeting a specific performance goal. The Division of Hospital Medicine adopted phlebotomy reduction as a performance metric. Resident champions introduced the program at monthly orientations and didactic sessions, emphasizing the patient experience related to phlebotomy.

Phlebotomy data were obtained from the electronic medical record (Epic Systems Corporation) and analyzed with R, version 3.1.3 (Foundation for Statistical Computing). The primary outcome was the monthly number of phlebotomies per patient per day, defined as the total number of patient phlebotomies divided by the sum of the lengths of stay for the patients on whom they were done. Patients in the intensive care unit, blood cultures, and perioperative collections were excluded from the analysis. Orders for blood culture were excluded because of technical issues preventing association of the order requisition with a specific blood draw. Perioperative collections were determined to be beyond the control of the Internal Medicine teams.

We used an interrupted time-series analysis to estimate changes in the use of phlebotomy in the postintervention period (January-October 2015) and control for preintervention trends. The preintervention period was defined as January 2013 to June 2014. The first 6 months after initiating the intervention were excluded to account for an educational lag. Institutional review board approval by the UCSF Committee on Human Research was not required for the intervention to reduce the number of needlesticks because it was deemed a quality improvement project that was overseen by the UCSF Division of Hospital Medicine Quality Improvement Committee.

Phlebotomy in the preintervention period (18 months) was done on 5527 patients vs 2635 patients in the postintervention period (10 months). The case mix index increased significantly in the postintervention compared with the preintervention period (mean, 1.38 vs 1.32, P < .001).

The average number of phlebotomies per patient per day in the preintervention period was 2.10 vs 1.78 in the postintervention period (P < .001). Our time-series analysis suggests that a secular trend toward a reduction of phlebotomies before the intervention (−0.008 needlesticks per month; P = .04) would not have led to the magnitude of the reduction in phlebotomies seen after the intervention (expected needlesticks at the end of October 2015, 1.95; actual, 1.75; relative reduction, −10.3%; 95% CI, −19.7% to −3.4%; P < .005) (Figure). After this one sustained decline, the trend toward improvement did not accelerate.

We present an approach to reducing laboratory use that focuses on the number of phlebotomies, and therefore on the importance of patient experience, with a simultaneous effort to provide high-value care. A patient-centered approach may provide physicians with a compelling reason to decrease unnecessary testing.

Our study has many limitations, including lack of randomization and lack of data about patient outcomes and satisfaction. The study probably also underestimated the overall phlebotomy burden through its exclusion of phlebotomies for blood culture and the possible need for multiple needle insertions to obtain a laboratory sample. The lack of acceleration in improvement after the intervention suggests that additional changes will be needed for continued gains.

Our program featured resident-physician champions and a system of timely, regular performance feedback with side-by-side team comparisons of phlebotomy rates. We used education and a financial incentive as additional tools to change the culture of inpatient laboratory testing. The transformation of practices in ordering laboratory services will require stronger systems-based changes.

Corresponding Author: Christopher Moriates, MD, Department of Medicine, University of California, San Francisco, 505 Parnassus Ave, Ste M1287, San Francisco, CA 94143 (CMoriates@medicine.ucsf.edu).

Published Online: April 4, 2016. doi:10.1001/jamainternmed.2016.0549.

Author Contributions: Ms Valencia had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Wheeler, Marcus, Nguyen, Kwong, Moriates.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Wheeler, Marcus, Nguyen, Moriates.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Kwong, Valencia.

Obtained funding: Nguyen.

Administrative, technical, or material support: All authors.

Study supervision: Wheeler, Moriates.

Conflict of Interest Disclosures: Dr Moriates reported having received royalties from McGraw-Hill for the textbook Understanding Value-Based Healthcare. No other disclosures were reported.

Funding/Support: This project was supported by a grant from the Lown Institute RightCare Alliance.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We acknowledge the rest of the Think Twice, Stick Once team without whom this study could not have been done, including Alvin Rajkomar, MD, Rachel Greenblatt, MD, Talia Kahn, MD, Kevin Yee, MD, David Margolius, MD, Hana Lim, MD, Chelsea Bowman, MD, Peggy Bui, MD, Nicole Kim, MD, and William Poe, MD, all from the Department of Medicine at the University of California, San Francisco. None of these contributors received any financial compensation for their work.