PROPOSAL TITLE: Establishing a Lab Stewardship Program to Reduce Lab Overutilization

PROJECT LEAD(S): Parul Bhargava, MD, Anoop Muniyappa, MD, MS, Teddy Peng, MD

EXECUTIVE SPONSOR(S): Amy Lu, MD, MPH, Andrew Auerbach­­­, MD

ABSTRACT 

When compared to peer academic medical centers, UCSF orders significantly more lab tests per inpatient hospitalization. Many lab tests are overutilized. To date, there is no Lab Stewardship Program at UCSF to assess lab utilization, review evidence, perform gap analyses, drive initiatives for safe and responsible lab stewardship, and engage in continuous improvement. Unnecessary labs contribute to unnecessary cost and wasted laboratory staff time, and can lead to excess venipuncture, iatrogenic anemia, prolonged length of stay, and increased mortality in hospitalized patients. Through analyzing UCSF lab tests by volume and cost, value-based care guidelines, and potential for successful clinical decision support-based interventions, our team has identified three lab tests as opportunities (CBC with differential, prealbumin, and free T4) for improvement with a high chance of success. We propose development of a Lab Stewardship Program led by the Medical Director of UCSF Labs to safely reduce unnecessary lab testing at UCSF starting with these 3 pilot tests. We aim to establish processes and protocols to implement and monitor improvement efforts, further refine our framework to identify opportunities to reduce unnecessary lab testing, and replicate these improvement processes for additional lab tests in the future. By focusing on EHR-based interventions and developing tools for real-time monitoring, we hope to create effective and sustainable lab reduction efforts.

TEAM 

• Parul Bhargava, MD (Senior Medical Director of UCSF Labs; project lead)
• Anoop Muniyappa, MD, MS (Hospital Medicine Faculty; informatics project co-lead)
• Theodore Peng, MD (Hospital Medicine Faculty; operations project co-lead)
• Janelle Lee, PhD (Senior Quality Improvement Programmer; project analyst)
• Brandon Scott, MD, MBA (Hospital Medicine Director of Value Improvement, UCSF Health Associate Medical Director for Value Improvement; project mentor)
• Sajan Patel, MD (Hospital Medicine Director of Quality Improvement; project mentor)
• Armond Esmaili, MD (Neurological Surgery Medical Director of Quality and Safety; project mentor)
• Andrew Auerbach, MD (Hospital Medicine Professor; executive sponsor)
• Amy Lu, MD, MPH (Chief Quality Officer; executive sponsor)

PROBLEM 

• Overall: Compared to peer academic medical centers (AMCs), UCSF orders significantly greater lab tests per inpatient hospitalization. Many lab tests are overutilized, leading to excess cost, staff time, and phlebotomy. See Appendix 1 for more details. UCSF currently does not have a comprehensive lab stewardship program to identify unnecessary lab testing. Through an analysis of UCSF lab tests by volume and cost, value-based care guidelines, and potential for successful clinical decision support (CDS)-based interventions, our team has identified three lab tests as opportunities for improvement with a high chance of success. Our goal is for the approach presented below to serve as the scaffolding for a comprehensive lab stewardship program at UCSF that will yield additional cost savings beyond this initial year.
• CBC with differential, prealbumin, and free T4 (thyroxine) are overutilized inpatient lab tests at UCSF that contribute to unnecessary cost, wasted laboratory staff time, and potentially excess venipuncture, and often do not align with clinical and value-based care guidelines. Guidelines and prior evidence suggest that many inpatient CBCs with differential (CBCD) can be converted to CBCs without differential (CBC), prealbumin ordering can be reduced, and free T4 testing can be done as a reflex to abnormal TSH without affecting patient quality or safety. In addition to provider education, these lab tests are well-suited for simple electronic health record (EHR) modifications, behavioral nudges, and CDS to effectively reduce unnecessary testing.
• Unnecessary lab tests also have negative effects on patients and lead to unnecessary additional costs. Studies show that routinely ordered lab tests can lead to iatrogenic anemia, increased need for blood transfusions, prolonged length of stay, and increased mortality in hospitalized patients.^1,2
• CBC with differential is one of the most frequently ordered inpatient lab tests, and is uniquely resource, time, and labor intensive. The UCSF clinical laboratory completed ~136,000 inpatient CBCDs in 2023. While the first pass of analyzing a WBC differential is automated, a significant proportion (~65%) require making a smear and manual counts, and a subset of these require clinicalpathologist review. As a result, there is a significant increase in reagent used, laboratory technologist time, and clinicalpathologist time to complete a CBCD when compared to a CBC. Additionally, a CBCD can take up to 4 times longer to complete and costs approximately $10 more in direct cost, which does not include the significant indirect costs mentioned above. See Appendix 2 for calculations.
• Many WBC differentials are unnecessary, and a significant proportion of CBC with differential can be safely converted to CBC alone. UCSF data shows that 6-7% of all CBCDs are repeated within 24 hours, and CBCDs are repeated even when the prior differentials are normal, both of which contribute to likely unnecessary testing. Published expert opinions recommenda time interval of 1.5 to up to 4 days for stable patients (excluding those on chemotherapy and being assessed for neutropenia).^3,4 Choosing Wisely also recommends against performing repetitive tests like CBCs at regular intervals in the face of clinical and lab stability.⁵
• Prealbumin is commonly ordered in the inpatient setting as a marker for poor nutrition, but this is not supported by clinical practice guidelines. The American Society for Parenteral and Enteral nutrition guidelines and the Choosing Wisely campaign recommend against using prealbumin as a marker for nutritional status or to screen or diagnose malnutrition.^6,7 The UCSF Clinical Laboratory completed nearly 3,000 inpatient prealbumin tests in 2023, many of which can be eliminated based on current guidelines.
• Free T4 is commonly co-ordered with TSH, which is not recommended by value-based care guidelines and results in unnecessary free T4 testing. The Choosing Wisely campaign recommends only a TSH for initial testing in suspected non-neoplastic thyroidal disease with follow up free T4 in only those with an abnormal TSH.^8,9 The UCSF Clinical Laboratory completed over 7,000 inpatient TSH tests in 2023. Approximately 80% of all inpatient TSH tests result in a normal TSH value and do not require a free T4, but over half of all TSH tests are initially co-ordered with a free T4 (~3,500). This means that approximately 60% (or 2,100) of free T4 tests could be avoided with a TSH that reflexes to a free T4 only when abnormal.
• Prior smaller-scale initiatives to measure and reduce unnecessary lab testing have been effective, but not sustained on a larger scale at UCSF and other AMCs. In the 2010s, UCSF saw several quality improvement initiatives to reduce unnecessary iCa, CBC, and electrolyte lab draws.^10,11 While successful, these were done prior to the implementation of an EHR. Now, with the EHR in place, this is a timely opportunity to make larger-scale, sustainable changes to improve lab stewardship. Additionally, the residents in the Neurosurgical Department at UCSF are currently participating in a REFLECT project dedicated to safely reducing unnecessary CBCs with differential, whether through conversion to CBC or through mindful cessation of excess lab draws. As Neurosurgery has the second largest patient volume at UCSF, this is a unique opportunity to collaborate for greater impact.

TARGET 

• Key Target Goal: Our goal is to establish a UCSF Lab Stewardship program, an evidence-based framework to identify overutilized lab tests, and a replicable process to safely reduce unnecessary lab testing at UCSF.
  • Our initial targets are to reduce inpatient WBC differential lab testing by 10-30% (~13,600-40,800), inpatient prealbumin lab testing by 60-80% (~1,800-2,400), and inpatient free T4 lab testing by 60% (~2,100) within 1 academic year.
  • We plan to measure the following inpatient key performance indicators over time: overall volume of CBCD, ratio of CBCD to CBC, volume of prealbumin, volume of free T4, number of each of these tests per inpatient admission, total lab tests per inpatient admission (which can be compared to benchmarks for similar AMCs), and estimated cost savings per admission.
  • We will also explore ways to measure the following additional areas where we anticipate reductions (reagent use, laboratory staff time, pathologist time, phlebotomy requests, iatrogenic anemia, clinical time reviewing lab tests) and improvements (patient experience scores).
  • Key Counterbalance Measures: Our aim is to ensure there is no increase in adverse events on involved service lines by reviewing key markers of safety related to our interventions. We also aim to quantify any disparities in ordering of the specified lab tests and ensure that disparities are not created or exacerbated by our intervention.

GAPS 

• System Issues:
• There is no comprehensive lab stewardship program at UCSF to assess lab utilization, review evidence, perform gap analyses, drive initiatives for safe and responsible lab stewardship, and engage in continuous improvement.
• Technological Gaps:
• Certain ordersets (including admission ordersets) are in place for different services that default to CBCD instead of CBC when a differential may not be indicated.
• There is no measure in place to prevent providers from ordering repeat CBCD daily or even more frequently when this is often not indicated.
• There is no backend automation or CDS to guide providers on appropriate CBCD ordering or when to decrease frequency of CBCD ordering.
• There is no CDS to guide providers away from inpatient prealbumin ordering.
• There is no reflex testing to perform a free T4 lab test only if the TSH is abnormal (which is present for other lab studies, such as Urinalysis with reflex to Urine Culture).
• There is no reporting tool that enables high-level stakeholders or individual service lines to understand volumes of lab tests ordered, identify opportunities for improvement, and to monitor the impact of these improvement efforts over time.
• Educational Gaps:
• There are no established criteria at UCSF (whether as EHR-based CDS or hospital wide resources) to help guide providers on clinically appropriate ordering intervals for CBCD, inpatient prealbumin testing, and testing free T4 only if TSH levels are abnormal.
• Providers are not aware of the additional resources consumed when performing a CBCD vs. CBC, the additional cost and time required, and how this can lead to lab delays.
• Service lines and providers are unaware of their own individual performance and where they may be opportunity for improved lab stewardship.

INTERVENTION 

• Pre-Intervention Data Collection and Analytics We will stratify ordering of the selected lab tests by inpatient service to understand the services that will be most impacted by our interventions. Additionally, we will review the volume and results from a subset of cases with a normal WBC to determine the potential utility and safety of a CBC with reflex to differential if the WBC is abnormal. We will review volume and results from a subset of cases with 2 consecutive normal differentials to determine the potential utility and safety of building CDS to suggest discontinuing daily differential orders after the second normal differential. Lastly, we will perform a manual chart review over 2 weeks of patients admitted to the Hospital Medicine and Neurosurgery services to determine what proportion of CBCD are appropriate for conversion to CBC to better predict the impact of this initiative.
• Stakeholder Engagement We plan to meet with key stakeholders (i.e., inpatient service directors) from Hematology, Nutritional Services, and Endocrinology to better understand lab recommendations and safety concerns, share our understanding of current evidence and best practices, and ensure alignment with how best to move forward together with a focus on reducing unnecessary lab testing. We will also briefly survey providers from different services (focusing on high-volume services and the above services given specific needs) on current practices for the ordering and use of CBCD, prealbumin, and TSH and free T4, and the potential impact of efforts to reduce ordering of these labs. We anticipate building carveouts for services like Hematology/Oncology, where more frequent differential ordering may be desired/indicated.  
• EHR Modification (see further below) We will submit multiple modification requests to the APeX Clinical Context Committee (AC3) to: remove CBCD from ordersets where clinically appropriate, limit the frequency of CBCD ordering to no more than q24h, build a BPA to suggest conversion of a daily CBCD order to a CBC after 2 consecutive normal WBC and differential results, build clinical guidance into the prealbumin order to reduce ordering, and build a TSH with reflex to free T4 order. We have already begun building order specifications and gathering additional data to guide these changes, and plan to have all proposed modifications submitited to AC3 prior to the start of FY25. 
• Education and Dissemination We will create a graphical abstract to circulate in the UCSF newsletter to provide evidence-based guidance and support to reduce CBCD, prealbumin, and free T4 ordering. Additionally, we will circulate clinical guidance and order modification summaries to service directors and division chiefs to distribute to inpatient providers. Lastly, we will do targeted presentations and education with divisions/services, including residents, identified as frequent utilizers of CBCD, prealbumin, and free T4.
• Monitoring We will develop a dashboard to track ordering of lab tests stratified by clinical service over time to identify opportunities for improvement, assess the impact of the proposed interventions, measure performance, and provide real-time feedback to clinical service lines.
• Launch System Wide Our goal is to create sustainable, system-wide changes to improve lab stewardship. In collaboration with our informatics experts, we have determined that making EHR changes for a single service line or by context would likely be technically challenging and overly complicated. We will focus our efforts on ensuring strong education and awareness across services, thorough pretesting of the proposed changes with our APeX colleagues, gathering feedback from patient safety leaders, ensuring appropriate measurement of counterbalance/safety measures, and continuous feedback from departmental stakeholders.
• Extensive Safety Review We will manually review charts for a subset of patients that would be affected by our proposed EHR modifications to assess for any potential adverse impacts on patient safety, quality, delays in diagnosis. Additionally, we will verify with inpatient Safety Directors that there has not been an increase in adverse events due to order reduction.
• Health Equity We will assess ordering of the specified lab tests by preferred language, sex, race-ethnicity, and payor to evaluate any systematic differences in ordering by these health equity-related variables. We will reassess this after deployment of our interventions to ensure that they do not create or exacerbate disparities.

PROPOSED EHR MODIFICATIONS 

• We will modify ordersets that currently default to CBCD to CBC (with option for differential) where clinically appropriate.
• We will modify the CBCD order to prevent ordering more frequently than q24h, with exceptions for certain services where clinically appropriate (e.g. Hematology/Oncology)
• We will build of a BPA to suggest ordering a CBC after 2 consecutive CBCDs.
• We will build an alert into the prealbumin order to recommend against ordering in the inpatient setting as a marker of nutritional status with supporting evidence/educational links.
• We will collaborate with Lab Medicine to build a TSH with reflex to free T4 order and set this as the default/highest on the list for free T4 and related search terms.

RETURN ON INVESTMENT (ROI) 

• Our estimated ROI from this initiative ranges from $161,104.00 - $439,974.00 per year in saved direct costs from these initial 3 lab tests alone. This does not take into account significant indirect cost savings for phlebotomy (including nursing support), pathologist time to review a subset of differentials, and clinician time to review results. It also does not account for additional future lab tests that would be identified and reduced through the creation of the Lab Stewardship Program. Of note, as reimbursement for inpatient laboratory testing is typically DRG-based for the entire admission, reduction in testing does not result in reduction of revenue. See Appendix 2 below for additional details.

SUSTAINABILITY 

• This initiative to reduce inappropriately ordered lab tests, including CBC with differential, prealbumin, and free T4 will be sustained through clinical decision support and order changes in APeX, educational campaigns, visible data tracking, and the establishment of a Lab Stewardship program.
  • The multiple APeX order changes and embedded CDS will persist beyond the funding year and be maintained through existing APeX infrastructure.
  • Widespread educational campaigns will help to inform and engage stakeholders across departments about the importance and relevance of these APeX changes. With the rollout of these proposed APeX changes and these dissemination efforts, we expect that the proposed changes will be accepted and become standard ordering practice within the project year. The Lab Stewardship program will continue to monitor the ordering of these tests and can provide additional feedback, education, and support as needed in future.
  • The reporting tools developed through this project will continue to be accessible and enable providers and floor units to monitor their performance beyond the project year.
  • If our efforts are successful in reducing unnecessary testing for the aforementioned labs, this can be applied to inpatient pediatrics, obstetrics, and certain outpatient labs. Most excitingly, these interventions can also be applied to countless other inappropriately order lab tests in the hospital, leading to future cost savings.
  • As described above, these operational changes to ordering labs for hospitalized patients would be long-term and sustainable.

BUDGET 

• Salary support for project co-leads for project implementation: $40,000-$45,000
• Development of educational materials, data analytics, and IT support: $5,000-$10,000