Decreasing daily blood work in hospitals: What works and what doesn't

Blood over-testing: impact, ethical issues and mitigating actions

Plenty of studies demonstrate that hospital-acquired anemia (HAA) can increase transfusion rates, mortality, morbidity and cause unnecessary patient burden, including additional length of hospital stay, sleep disruption and venipuncture harms resulting from blood samples unlikely to change clinical management. Beyond patient costs, community costs should also be considered, such as laboratory time and resources waste, environmental impact, increasing pressure on labs and fewer tests available on time for patients who can benefit from them most. Blood over-testing does not support the principles of non-maleficence, justice and respect for patient autonomy, at the expense dubious beneficence. Reducing the number and frequency of orders is possible, to a certain extent, by adopting nudge strategies and raising awareness among prescribing doctors. However, reducing the orders may appear unsafe to doctors and patients. Therefore, reducing blood volume from each order is a better alternative, which is worth implementing through technological, purchasing and organizational arrangements, possibly combined according to need (smaller tubes, adequate analytic platforms, blind dilution, blood conservative devices, aggregating tests and laboratory units).

Successful demand management in diagnostic immunology testing

AIMS

We investigated whether we could have a material and sustained impact on immunology test ordering by primary care clinicians by building evidence-based and explanatory algorithms into test ordering software.

METHODS

A service evaluation revealed cases of over-requesting of antinuclear antibody, allergen-specific IgE and total IgE tests, and under-requesting of urine protein electrophoresis. We conducted a quality improvement programme to address this. We determined the most effective and efficient intervention would be to embed evidence-based and advice-based decision-support algorithms in the ordering software. Consultation with general practitioners revealed lack of knowledge and confidence about testing, and an appetite for support. We iteratively designed and implemented algorithms for the four sets of tests for the primary care practices in our catchment and made them available to other hospital trusts in our region. The ordering system now contains links to advice sheets for clinicians and their patients and to an email address for queries to the lab.

RESULTS

We observe large (36% to 88%) reductions in testing activity (workload) for the over-requested tests and large (28%-135%) increases for the under-requested test. We show that these changes are sustained. There have been no complaints from the clinicians and queries to the lab are now minimal (less than one per month on average).

CONCLUSIONS

Embedding algorithms in the ordering software can be acceptable to clinicians and have a major and sustained impact on overuse or underuse of tests. The algorithms can be replicated by other hospital trusts.

Frequency of Repeating Antinuclear Antibody Testing: When Less Is More

Objectives Antinuclear antibodies (ANA) are autoantibodies that are associated with and ordered to diagnose autoimmune connective tissue disease. ANA have high sensitivity (~98%) but low specificity (~75%), and because they can be found in healthy individuals and non-rheumatologic conditions leading to their elevation, ANA tests are often requested and interpreted inappropriately by clinicians. The aim of this study was to retrospectively assess how frequently ANA testing is repeated in the adult population of Saudi Arabia (SA) and which factors are associated with and lead to inappropriate testing. Methodology We investigated a study group of 40,634 adult patients who underwent 229,825 ANA tests from 2018 to 2022 in an academic hospital in Jeddah, SA. We took a random sample of 500 patients from the study group, along with their 998 ANA tests, to look in depth into our research questions. Variables related to patients, ANA tests, and ordering physicians were collected. Descriptive and analytical statistics were employed to address the research questions, and a p-value < 0.05 was considered statistically significant. Results We found 57% of the ordered ANA tests to have positive results, with the most common titers of mild positivity being 1:80 and 1:160. Most repeated ANA tests were ordered with an interval of more than one year, and when repeated, 67% of test results remained unchanged. The majority of seroconversions resulted from negative ANA tests or those with weak (titer 1:40) or mild positivity (titer 1:80-1:160). The results of the moderate (titer 1:320-1:640) and strong (titer ≥1280) positivity ANA tests did not change. Only 11% of repeated ANA tests were found to be appropriate for repetition. The most common specialties associated with ordering ANA tests in general were internal medicine, followed by rheumatology, and finally family medicine. Our correlation analysis revealed that being female, having systemic connective tissue disease, and having a rheumatologist as a specialist were all associated with ordering more than 10 ANA tests (p < 0.05). Conclusion Because the results of repeated ANA tests did not change much, our study suggests that the cost of repeating ANA tests and the subsequent potentially unnecessary interventions should all be carefully examined before scheduling a repeated ANA test. Further studies involving patients from SA and across wider healthcare settings (academic, community, and private hospitals and healthcare centers) are warranted.

Factors of a physician quality improvement leadership coalition that influence physician behaviour: a mixed methods study

BACKGROUND

A coalition (Strategic Clinical Improvement Committee), with a mandate to promote physician quality improvement (QI) involvement, identified hospital laboratory test overuse as a priority. The coalition developed and supported the spread of a multicomponent initiative about reducing repetitive laboratory testing and blood urea nitrogen (BUN) ordering across one Canadian province. This study's purpose was to identify coalition factors enabling medicine and emergency department (ED) physicians to lead, participate and influence appropriate BUN test ordering.

METHODS

Using sequential explanatory mixed methods, intervention components were grouped as person focused or system focused. Quantitative phase/analyses included: monthly total and average of the BUN test for six hospitals (medicine programme and two EDs) were compared pre initiative and post initiative; a cost avoidance calculation and an interrupted time series analysis were performed (participants were divided into two groups: high (>50%) and low (<50%) BUN test reduction based on these findings). Qualitative phase/analyses included: structured virtual interviews with 12 physicians/participants; a content analysis aligned to the Theoretical Domains Framework and the Behaviour Change Wheel. Quotes from participants representing high and low groups were integrated into a joint display.

RESULTS

Monthly BUN test ordering was significantly reduced in 5 of 6 participating hospital medicine programmes and in both EDs (33% to 76%), resulting in monthly cost avoidance (CAN$900-CAN$7285). Physicians had similar perceptions of the coalition's characteristics enabling their QI involvement and the factors influencing BUN test reduction.

CONCLUSIONS

To enable physician confidence to lead and participate, the coalition used the following: a simply designed QI initiative, partnership with a coalition physician leader and/or member; credibility and mentorship; support personnel; QI education and hands-on training; minimal physician effort; and no clinical workflow disruption. Implementing person-focused and system-focused intervention components, and communication from a trusted local physician-who shared data, physician QI initiative role/contribution and responsibility, best practices, and past project successes-were factors influencing appropriate BUN test ordering.

Characteristics promoting behaviour change: physician experience with a coalition-led quality improvement initiative to reduce excessive laboratory test ordering

BACKGROUND

Attempts have been made to reduce excessive laboratory test ordering; however, the problem persists and barriers to physician involvement in quality improvement (QI) remain. We sought to understand physician participation experience following a laboratory test overuse initiative supported by a QI coalition.

METHODS

As part of a larger mixed-methods study, structured virtual interviews were conducted with 12 physicians. The Theoretical Domains Framework (TDF) and the Behavioural Change Wheel (BCW) were used to identify characteristics that influence physician behaviour change for QI leadership and participation and appropriate blood urea nitrogen (BUN) test ordering. A content analysis of physicians' statements to the TDF was performed, resulting in overarching themes; relevant TDF domains were mapped to the intervention functions of the BCW.

RESULTS

Nine overarching themes emerged from the data. Eight of 14 TDF domains influence QI leadership and participation, and 10 influence appropriate BUN-test ordering behaviours. The characteristics participants described that promoted a change in their QI participation, leadership and appropriate BUN-test ordering were: QI education with hands-on training; physician peer mentorship/support; personnel assistance (QI and analytics) and communication from a trusted/credible physician leader who shares data and insights about the physician role in the initiative, clinical best practice and past project success. Other elements included: a simply designed initiative requiring minimal effort and no clinical workflow disruptions; revised order forms/panels and limiting test-order frequency when laboratory tests are normal. Additionally, various future intervention strategies were identified. For their initial initiative participation, physicians acknowledged coalition leader or member credibility was more important than awareness of the coalition.

CONCLUSIONS

Based on physicians' described perceptions and experiences, coalition characteristics that influenced their QI leadership and participation, and appropriate BUN-test ordering behaviours were revealed; these characteristics aligned to several TDF domains. The findings suggest that these behaviours are multidimensional, requiring a multistrategy approach to change behaviour.

Reduction of urea test ordering in the emergency department: multicomponent intervention including education, electronic ordering, and data feedback

Introduction

In the emergency department (ED), laboratory testing accounts for a significant portion of the medical assessment. Although excess laboratory test ordering has been proven to be prevalent, different types of interventions have been used to encourage a behavioural change in how physicians order tests. In one western Canadian hospital medicine program, a quality improvement project aimed to reduce the total monthly blood urea nitrogen (BUN) test ordered by physicians was found to be successful. The objective of this project was to evaluate a similar multicomponent intervention aimed at ED physician ordering, with the primary goal of reducing the number of monthly BUN tests ordered per ED visit.

Methods

A pre post intervention design was conducted over 12-months. The first intervention component was an educational presentation conducted by physician leaders. Second, a regularly used order panel within the ED electronic order system was modified, removing the BUN test. The third component involved audit and feedback; the total monthly BUN test ordered for the ED department post intervention start was shared with all ED physicians twice (at 5 and 12 months).An interrupted time series analysis was completed to evaluate the multicomponent intervention effect.

Results

The total monthly ordered BUN test declined from an average of 1905 pre-intervention to 448 post-intervention, and the total monthly BUN test to total ED visit ratio declined from 0.46 to 0.1. These results were a statistically significant reduction in physician BUN test ordering.

Conclusions

Targeted education, order panel design and data feedback interventions can impact physician ordering behaviour in the emergent healthcare context, where diagnostic tests are often over used.

Supplementary Information

The online version contains supplementary material available at 10.1007/s43678-022-00333-w.

Association Between Health System Factors and Utilization of Routine Laboratory Tests in Clinical Teaching Units: a Cohort Analysis

BACKGROUND

Few studies have looked at health system factors associated with laboratory test use.

OBJECTIVE

To determine the association between health system factors and routine laboratory test use in medical inpatients.

DESIGN

We conducted a retrospective cohort study on adult patients admitted to clinical teaching units over a 3-year period (January 2015 to December 2017) at three tertiary care hospitals in Calgary, Alberta.

PARTICIPANTS

Patients were assigned to a Case Mix Group+ (CMG+) category based on their clinical characteristics, and patients in the top 10 CMG+ groups were included in the cohort.

EXPOSURES

The examined health system factors were (1) number of primary attending physicians seen by a patient, (2) number of attending medical teams seen by a patient, (3) structure of the medical team, and (4) day of the week.

MAIN MEASURES

The primary outcome was the total number of routine laboratory tests ordered on a patient during their admission. Statistical models were adjusted for age, sex, length of stay, Charlson comorbidity index, and CMG+ group.

RESULTS

The final cohort consisting of 36,667 patient-days in hospital (mean (SD) age 62.5 (18.4) years) represented 5071 unique hospitalizations and 4324 unique patients. Routine laboratory test use was increased when patients saw multiple attending physicians; with an adjusted incidence rate ratio (IRR) of 1.46 (95% CI, 1.37-1.55) for two attending physicians, and 2.50 (95% CI, 2.23-2.79) for three or more attending physicians compared to a single attending physician. The number of routine laboratory tests was slightly lower on weekends (IRR 0.98, 95% CI, 0.96-0.99) and on teams without a senior resident as part of their team structure (IRR 0.89, 95% CI 0.830.96).

CONCLUSIONS

The associations observed in this study suggest that breaks in continuity of care, including increased frequency in patient transfer of care, may impact the utilization of routine laboratory tests.

Predicting abnormal laboratory blood test results in the intensive care unit using novel features based on information theory and historical conditional probability: Observational Study (Preprint)

Background

Redundancy in laboratory blood tests is common in intensive care units (ICUs), affecting patients’ health and increasing health care expenses. Medical communities have made recommendations to order laboratory tests more judiciously. Wise selection can rely on modern data-driven approaches that have been shown to help identify low-yield laboratory blood tests in ICUs. However, although conditional entropy and conditional probability distribution have shown the potential to measure the uncertainty of yielding an abnormal test, no previous studies have adapted these techniques to include them in machine learning models for predicting abnormal laboratory test results.

Objective

This study aimed to address the limitations of previous reports by adapting conditional entropy and conditional probability to extract features for predicting abnormal laboratory blood test results.

Methods

We used an ICU data set collected across Alberta, Canada, which included 55,689 ICU admissions from 48,672 patients. We investigated the features of conditional entropy and conditional probability by comparing the performances of 2 machine learning approaches for predicting normal and abnormal results for 18 blood laboratory tests. Approach 1 used patients’ vitals, age, sex, and admission diagnosis as features. Approach 2 used the same features plus the new conditional entropy–based and conditional probability–based features. Both approaches used 4 different machine learning models (fuzzy model, logistic regression, random forest, and gradient boosting trees) and 10 metrics (sensitivity, specificity, accuracy, precision, negative predictive value [NPV], F₁ score, area under the curve [AUC], precision-recall AUC, mean G, and index balanced accuracy) to assess the performance of the approaches.

Results

Approach 1 achieved an average AUC of 0.86 for all 18 laboratory tests across the 4 models (sensitivity 78%, specificity 84%, precision 82%, NPV 75%, F₁ score 79%, and mean G 81%), whereas approach 2 achieved an average AUC of 0.89 (sensitivity 84%, specificity 84%, precision 83%, NPV 81%, F₁ score 83%, and mean G 84%). We found that the inclusion of the new features resulted in significant differences for most of the metrics in favor of approach 2. Sensitivity significantly improved for 8 and 15 laboratory tests across the different classifiers (minimum P<.001 and maximum P=.04). Mean G and index balanced accuracy, which are balanced performance metrics, also improved significantly across the classifiers for 6 to 10 and 6 to 11 laboratory tests. The most relevant feature was the pretest probability feature, which is the probability that a test result was normal when a certain number of consecutive prior tests was already normal.

Conclusions

The findings suggest that conditional entropy–based features and pretest probability improve the capacity to discriminate between normal and abnormal laboratory test results. Detecting the next laboratory test result is an intermediate step toward developing guidelines for reducing overtesting in the ICU.

Reducing Unnecessary Laboratory Utilization in the Medical ICU: A Fellow-Driven Quality Improvement Initiative

OBJECTIVES:

Overutilization of laboratory services is now recognized as harmful to patients and wasteful. In fact, the American Board of Internal Medicine’s Choosing Wisely campaign recommends against ordering routine testing that does not answer a clinical question. Per peer benchmarking, our institution as a whole occupied an extreme outlier position at the 100th percentile for laboratory utilization. We sought to address this problem starting in our medical ICUs with a quality improvement project.

DESIGN:

Quality improvement project using the design, measure, analyze, improve, and control process. The primary endpoint was a sustained reduction in laboratory utilization. Counterbalance metrics were also followed, and these included mortality, renal replacement therapy initiation rates, stat laboratory orders, and central catheter–associated blood stream infections.

SETTING:

The medical ICU at the Ohio State University Medical Center.

PATIENTS:

All patients admitted to the medical ICU from March 2019 to March 2020.

INTERVENTIONS:

Root causes were identified and addressed with the implementation of a wide range of interventions involving a multidisciplinary team led by trainee physicians.

MEASUREMENTS AND MAIN RESULTS:

There was a sustained 20% reduction in the number of tests performed per patient day, with no change in the counterbalance metrics.

CONCLUSIONS:

Trainees can affect positive change in the culture and processes at their institutions to safely reduce laboratory utilization.

Reducing inappropriate blood testing in haematology inpatients: A multicentre quality improvement project

Haematology inpatients are subject to extensive blood testing and many of these tests could be deemed inappropriate as they are not indicated for monitoring or clinical symptoms. Unnecessary testing exposes the patient to the risks of phlebotomy and adds resources' strain to the NHS.Our aim was to reduce the number of inappropriate blood tests performed on haematology inpatient wards.Quality improvement projects (QIPs) were performed in four haematology units introducing inpatient blood testing schedules (BTS) or providing staff education on current schedules.A reduction in inappropriate or overall blood testing was achieved at every site where a BTS was implemented, with a median reduction in inappropriate blood testing of 24.7% and estimated cost savings of up to £38,438 per annum.This QIP can be safely adapted to a variety of inpatient settings and is associated with cost savings. This initiative could be extended to other inpatient departments throughout the NHS.

Potentially redundant repeat liver function test ordering practices in australian hospitals: A 5-year multicentre retrospective observational study

BACKGROUND

Repeat Liver Function Tests (LFTs) are often necessary for monitoring purposes, but retesting within a short time interval may suggest potentially redundant repeat test (PRRT) ordering practices. We aimed to determine the proportion of potentially redundant repeat LFT ordering and identify associated factors in hospitals.

METHODS

A 5-year (2014-2018) retrospective cohort study in six hospitals in New South Wales, Australia. A total of 131 885 patient admissions with repeat LFTs in the general ward (n = 102 852) and intensive care unit (ICU) (n = 29 033) met the inclusion criteria. Existing guidelines do not support retesting LFT for at least 48-72 hours. We used 24 hours as a conservative minimum retesting interval to examine PRRT ordering. We fit binary logistic regression to identify factors associated with PRRT ordering in two conditions with the highest repeat LFTs.

RESULTS

There were a total of 298 567 repeat LFTs (medians of 2 repeats/admission and retesting interval of 25.6 hours) in the general ward and 205 929 (medians of 4 repeats/admission and retesting interval of 24.1 hours) in the ICU. The proportions of PRRT ordering were 35.2% (105 227/298 567) and 47.7% (98 307/205 929) in the general ward and ICU, respectively. The proportions of patients who received at least one PRRT were 52.3% (53 766/102 852) and 83.9% (24 365/29 033) in the general ward and ICU, respectively. Age, gender and the number of comorbidities and procedures were associated with the likelihood of ordering PRRT depending on the settings.

CONCLUSION

Repeat LFT testing is common in Australian hospitals, often within 24 hours, despite guidelines not supporting too-early repeat testing. Further research should be conducted to understand whether better adherence to existing guidelines is required, or if there is any case for guidelines to be updated based on certain patient subpopulations.

Duplicate Testing: Enhancing Transitions in Care Communication in the Infusion Center and Emergency Department Settings

Two objectives of the American Board of Internal Medicine's Choosing Wisely® initiative include avoiding duplicate testing and choosing care that is free from harm. Oncology nurses in an academic comprehensive cancer center observed a pattern of testing duplication and related dissatisfaction among providers and patients. A quality improvement project was initiated to quantify the problem and reduce duplication by implementing collaborative solutions to improve interdepartmental communication during handoffs. Postimplementation data revealed a 35% decrease of testing duplication in the patient population.

Inappropriate laboratory testing in internal medicine inpatients: Prevalence, causes and interventions

BACKGROUND

To reduce overutilization of laboratory testing many interventions have been tried, but selecting the most effective intervention for a given setting is challenging. To be sustainable, interventions need to align with healthcare providers' needs and daily practices. This study aimed to assess the extent of overutilization and the perspectives of healthcare providers, which may be used to guide the choice of intervention.

METHODS

The extent of inappropriate laboratory testing in internal medicine inpatients was evaluated using a database. Surveys and focus groups were used to investigate healthcare providers' perceptions on its causes and solutions.

RESULTS

On average, patients had 5.7 laboratory orders done during the first week of admission, whereas guidelines advise performing laboratory testing no more than twice per week. Repeat testing of normal test results occurred in up to 85% of patients. The frequency of laboratory testing was underestimated by survey responders, even though the majority of responders (78%) thought that laboratory tests are ordered too frequently. Residents were considered to be most responsible for laboratory test ordering.The primary causes of overutilization discussed were personal factors, such as a lack of awareness and knowledge, as well as feelings of insecurity. Regarding possible solutions, residents generally recommended educational interventions, whereas specialists tended to favour technical solutions such as lockouts.

CONCLUSION

Inappropriate laboratory testing is common in internal medicine. The most important causes are a lack of awareness and knowledge, especially in residents. The intervention most favoured by residents is education, suggesting educational interventions may be most applicable.