Reducing unnecessary testing: an intervention to improve resident ordering practices

The preparedness and knowledge of pharmacists and general practitioners in managing human monkeypox: a highly spreading infectious disease

BACKGROUND

After the era of the COVID-19 pandemic, the role of pharmacists was emphasized in the battle against highly spreading and infectious diseases like human Monkeypox (hMPV).

AIM

Assess the hMPV knowledge of the community, clinical pharmacists, and general practitioners (GPs) and raise their awareness about hMPV.

METHODS

A web-based questionnaire was distributed randomly to Egyptian community pharmacists, clinical pharmacists, and GPs from all governorates. The questionnaire was divided into two sections: one for demographic information and the other for hMPV knowledge (nature of the disease, incubation period, transmission, symptoms, Prophylaxis, Prevention, and management). The evidence-based answers were provided after completing the submission. Data were descriptively analyzed using IBM SPSS software.

RESULTS

From a total of 753 respondents, only 710 participants were included in the final data analysis. The % of respondents who presented good total knowledge scores about hMPV was comparable between study groups (P = 0.826). There were no differences between groups identifying different disease clinical characteristics (P = 0.689) and hMPV management (P = 0.324). Community pharmacists had better knowledge scores than GPs in the prevention and prophylaxis domain (P = 0.037).

CONCLUSION

Pharmacists and GPs have good and similar knowledge levels of hMPV. However, a gap exists in recognizing the right hMPV incubation period, prophylaxis, and omitting antibiotics from hMPV management. Pharmacists and GPs are the frontline health care providers (HCPs), so they would require more knowledge enhancement about such contagious diseases to offer the best possible patient care.

Evaluation of human monkeypox knowledge and beliefs regarding emerging viral infections among healthcare workers

OBJECTIVES

The purpose of this study was to evaluate possible factors that might be accompanied by high level of human monkey pox (HMPX) knowledge and to explain the relationship between HMPX knowledge and Beliefs regarding emerging viral infections.

STUDY DESIGN

A descriptive cross-sectional study was conducted for the implementation of this study.

METHODS

Study was conducted at two general hospitals in Mansoura City (Old General Hospital and International Hospital) El Dakahlia Governorate among 620 healthcare workers (HCWs) using a self-managed questionnaire for 1 week (1 to 7 January 2023). The questionnaire has items adapted from the previously published literature to assess HMPX knowledge and Beliefs regarding emerging viral infections.

RESULTS

The mean age of the study sample was 27.97 years and most of them were female (86.1%). Physicians and other HCWs (nurses, laboratory technicians, radiographer technicians, and pharmacists) had significantly different levels of knowledge of monkeypox for the majority of the questions. A higher belief was found among two items: viruses are biological weapons manufactured by the superpowers to take global control and the government is misleading the public about the cause of the virus.

CONCLUSION

This study discovered lower levels of knowledge of HMPX among HCWs in Egypt. Beliefs about emerging viral infections were widespread, and future research should look into their potential negative impact on health behavior.

Reducing Unnecessary Complete Blood Count Ordering Through Education and Standardization: A Quality Improvement Initiative

BACKGROUND AND OBJECTIVES

The American Board of Internal Medicine's Choosing Wisely campaign recommends against ordering repetitive complete blood counts (CBC) in the face of clinical and laboratory stability.

METHODS

Consecutive patients admitted to a teaching team were included. Intervention 1 was an educational lecture outlining costs of and indications for CBC ordering. Intervention 2 added a simplified algorithm to help providers determine the need for a daily CBC. The primary outcome measure was the number of CBCs ordered per number of patients per day. The secondary outcome measure was net cost saved. The process measures were lecture/poster and algorithm utilization rates. The balancing measure was emergency department visits/readmissions within 7 days of discharge. A statistical process control chart was generated to assess special cause variation. Using R software version 3.5.2, a 2-sample t test and Fisher exact test differences between groups in the outcome and balancing measures.

RESULTS

One hundred ten patients were included over a 62-day period. The difference between the pre-intervention group and both interventions combined was significant ( P = .000317). Special cause variation was observed after institution of both interventions in conjunction. Net costs saved totaled $43 482. Emergency department visits/readmissions within 7 days were similar between the groups ( P = .1403).

CONCLUSIONS

Complete blood count ordering patterns and costs were improved through education and providing a decision support tool in the form of a simplified algorithm, without increasing 7-day emergency department visits/readmissions. The algorithm, far less detailed than that previously published, still resulted in significant improvement without unintended consequences, making for a safe and potentially sustainable intervention.

Interventions to reduce repetitive ordering of low-value inpatient laboratory tests: a systematic review

BACKGROUND

Over-ordering of daily laboratory tests adversely affects patient care through hospital-acquired anaemia, patient discomfort, burden on front-line staff and unnecessary downstream testing. This remains a prevalent issue despite the 2013 Choosing Wisely recommendation to minimise unnecessary daily labs. We conducted a systematic review of the literature to identify interventions targeting unnecessary laboratory testing.

METHODS

We systematically searched MEDLINE, EMBASE, Cochrane Central and SCOPUS databases to identify interventions focused on reducing daily complete blood count, complete metabolic panel and basic metabolic panel labs. We defined interventions as 'effective' if a statistically significant reduction was attained and 'highly effective' if a reduction of ≥25% was attained.

RESULTS

The search yielded 5646 studies with 41 articles that met inclusion criteria. We grouped interventions into one or more categories: audit and feedback, cost display, education, electronic medical record (EMR) change, and policy change. Most interventions lasted less than a year and used a multipronged approach. All five strategies were effective in most studies with EMR change being the most commonly used independent strategy. EMR change and policy change were the strategies most frequently reported as effective. EMR change was the strategy most frequently reported as highly effective.

CONCLUSION

Our analysis identified five categories of interventions targeting daily laboratory testing. All categories were effective in most studies, with EMR change being most frequently highly effective.

PROSPERO REGISTRATION NUMBER

CRD42021254076.

Monkeypox Knowledge and Confidence in Diagnosis and Management with Evaluation of Emerging Virus Infection Conspiracies among Health Professionals in Kuwait

As the 2022 human monkeypox (HMPX) multi-country outbreak is spreading, the response of healthcare workers (HCWs) is central to mitigation efforts. The current study aimed to evaluate HMPX knowledge and confidence in diagnosis and management among HCWs in Kuwait. We used a self-administered questionnaire distributed in July–August 2022 through a snowball sampling approach. The survey items evaluated HMPX knowledge, confidence in diagnosis and management of the disease, and the belief in conspiracies regarding emerging virus infections (EVIs). The sample size was 896 HCWs: nurses (n = 485, 54.1%), pharmacists (n = 154, 17.2%), physicians (n = 108, 12.1%), medical technicians/allied health professionals (MT/AHP, n = 96, 10.7%), and dentists (n = 53, 5.9%). An overall low level of HMPX knowledge was noticed for items assessing virus transmission and non-cutaneous symptoms of the disease, with higher knowledge among physicians. Approximately one-fifth of the study sample agreed with the false notion that HMPX is exclusive to male homosexuals (n = 183, 20.4%), which was associated with lower knowledge with higher frequency among MT/AHP compared to nurses, physicians, and pharmacists. Confidence levels were low: confidence in diagnosis based on diagnostic tests (n = 449, 50.1%), confidence in the ability to manage the HMPX (n = 426, 47.5%), and confidence in the ability to diagnose HMPX clinically (n = 289, 32.3%). Higher confidence levels were found among nurses and participants with postgraduate degrees. Higher embrace of conspiracy beliefs regarding EVIs was noticed among participants with lower knowledge, and among those who agreed or were neutral/had no opinion regarding the false idea of HMPX exclusive occurrence among male homosexuals, while lower levels of belief in conspiracies were noticed among physicians, dentists, and pharmacists compared to MT/AHP. Variable levels of HMPX knowledge were observed in this study per item, with low level of knowledge regarding virus transmission. Differences in knowledge and confidence levels in diagnosis and management of HMPX should be considered in education and training aiming to prepare for outbreak response. The relatively high prevalence of embracing conspiratorial beliefs regarding EVIs is worrisome and needs proper interventions. The attitude towards male homosexuals’ role in monkeypox spread should be evaluated in future studies considering the possibility of stigma and discrimination in this most-at-risk group.

Sustainable Laboratory-Driven Method to Decrease Repeat, Same-Day WBC Differentials at a Tertiary Care Center

OBJECTIVES

A CBC with WBC differential is often ordered when a CBC alone would be sufficient for patient care. Performing unnecessary WBC differentials adds to costs in the laboratory. Our objective was to implement a laboratory middleware algorithm to cancel repeat, same-day WBC differentials to achieve lasting improvements in laboratory resource allocation.

METHODS

Repeat same-day WBC differentials were first canceled only on intensive care unit samples; after a successful trial period, the algorithm was applied hospital-wide. We retrospectively reviewed CBC with differential orders from pre- and postimplementation periods to estimate the reduction in WBC differentials and potential cost savings.

RESULTS

The algorithm led to a monthly WBC differential cancellation rate of 5.40% for a total of 10,195 canceled WBC differentials during the cumulative postimplementation period (September 25, 2019, to December 31, 2020). Nearly all (99.94%) differentials remained canceled. Most patients only had one WBC differential canceled (range, 1-38). Savings estimates showed savings of $0.99 CAD per canceled differential and 1,060 minutes (17.7 hours) of technologist time.

CONCLUSIONS

A middleware algorithm to cancel repeat, same-day WBC differentials is a simple and sustainable way to achieve lasting improvements in laboratory utilization.

Cost of Care in Open Cystectomy Patients Across Time and Space: Does it matter?

BACKGROUND: Many variables may affect the cost of open radical cystectomy (RC) care, including surgical approach, diversion type, patient comorbidities, and postoperative complications. OBJECTIVE: To determine factors associated with changes in cost of care following open radical cystectomy (ORC) for bladder cancer using the National Inpatient Sample (NIS). METHODS: Patients in the NIS with a diagnosis of bladder cancer who underwent ORC with ileal conduit from 2012–2017 using ICD-9-CM and ICD-10-CM codes were identified. Baseline demographics including age, race, region, postoperative complications, and length of stay were obtained. Univariable and multivariable logistic regression were used to identify factors associated with cost variation including demographics, clinical characteristics, surgical factors, and discharge quarter (Q1-Q4). RESULTS: 5,189 patients were included in the analysis, with 4,379 at urban teaching hospitals. On multivariable regression analysis, female sex [$1,734 ($1,024–2,444) p < 0.001)], a greater Elixhauser comorbidity score [$93 ($62–124), p < 0.001], presence of any inpatient complication [$1,531 ($894–2,168), p < 0.001], and greater length of stay [$1,665 ($1,536–1,793), p < 0.001] were associated with a greater cost of hospitalization. Discharge in Q3 (July to September) relative to Q2 (April to June) was associated with a higher cost [$1,113 ($292–1,933), p = 0.008. Trends were similar at urban non-teaching and rural hospitals, except discharge quarter was not associated with a significant change in cost. CONCLUSIONS: Significant differences in cost of ORC with ileal conduit exist with respect to patient sex, medical comorbidities, and discharge timing. These differences may relate to greater disease burden in female patients, patient complexity, and variation in postoperative care in academic programs.

Improving the Timing of Laboratory Studies in Hospitalized Children: A Quality Improvement Study

OBJECTIVES

For hospitalized children and their families, laboratory study collection at night and in the early morning interrupts sleep and increases the stress of a hospitalization. To change this practice, our quality improvement (QI) study developed a rounding checklist aimed at increasing the percentage of routine laboratory studies ordered for and collected after 7 am.

METHODS

Our QI study was conducted on the pediatric hospital medicine service at a single-site urban children's hospital over 28 months. Medical records from 420 randomly selected pediatric inpatients were abstracted, and 5 plan-do-study-act cycles were implemented during the intervention. Outcome measures included the percentage of routine laboratory studies ordered for and collected after 7 am. The process measure was use of the rounding checklist. Run charts were used for analysis.

RESULTS

The percentage of laboratory studies ordered for after 7 am increased from a baseline median of 25.8% to a postintervention median of 75.0%, exceeding our goal of 50% and revealing special cause variation. In addition, the percentage of laboratory studies collected after 7 am increased from a baseline median of 37.1% to 76.4% post intervention, with special cause variation observed.

CONCLUSIONS

By implementing a rounding checklist, our QI study successfully increased the percentage of laboratory studies ordered for and collected after 7 am and could serve as a model for other health care systems to impact provider ordering practices and behavior. In future initiatives, investigators should evaluate the effects of similar interventions on caregiver and provider perceptions of patient- and family-centeredness, satisfaction, and the quality of patient care.

Taking the Leap Toward Cost-Conscious Education in Obstetrics and Gynaecology: A Preliminary Randomized Controlled Trial

OBJECTIVE

Residents have a professional obligation with respect to the stewardship of health care resources, yet there is a paucity of research on how to improve residents' cost-awareness. Rising health care expenditures highlight a critical need to improve education related to this competency. This study aimed to test if an educational module can teach residents to make cost-conscious decisions and reduce health care spending.

METHODS

All Canadian obstetrics and gynaecology residents in 2017 were eligible to participate in this randomized controlled trial. The study was administered online via REDCap. Interested residents were enrolled, stratified by level of training, and block randomized. Residents completed a survey to determine their management of 4 obstetrical scenarios. The intervention group reviewed an educational module on cost-effective ordering prior to completing the survey; the control group was given the option to review the module afterward. The primary outcome was mean total expenditures, compared between the 2 groups using the t test.

RESULTS

Eighty-five residents were enrolled between August and November 2017, and 63 residents from 13 Canadian residency programs completed the study requirements (33 control and 30 intervention). Mean total expenditure was CAD$291.03 (95% CI 259.38-322.68) versus CAD$192.98 (95% CI 170.67-215.29) for the control and intervention groups, respectively. These figures corresponded to a 33.69% or CAD$98.05 reduction in total expenditures (P = 0.0001).

CONCLUSION

This educational module decreased expenditures by Canadian obstetrics and gynaecology residents managing hypothetical obstetrical cases. This introduces a potential curriculum innovation to improve resident education in judicious use of health care resources.

Developing and maintaining clinical decision support using clinical knowledge and machine learning: the case of order sets

Development and maintenance of order sets is a knowledge-intensive task for off-the-shelf machine-learning algorithms alone. We hypothesize that integrating clinical knowledge with machine learning can facilitate effective development and maintenance of order sets while promoting best practices in ordering. To this end, we simulated the revision of an "AM Lab Order Set" under 6 revision approaches. Revisions included changes in the order set content or default settings through 1) population statistics, 2) individualized prediction using machine learning, and 3) clinical knowledge. Revision criteria were determined using electronic health record (EHR) data from 2014 to 2015. Each revision's clinical appropriateness, workload from using the order set, and generalizability across time were evaluated using EHR data from 2016 and 2017. Our results suggest a potential order set revision approach that jointly leverages clinical knowledge and machine learning to improve usability while updating contents based on latest clinical knowledge and best practices.

Expert Recommendations on Frequency of Utilization of Common Laboratory Tests in Medical Inpatients: a Canadian Consensus Study

BACKGROUND

Repetitive inpatient laboratory testing in the face of clinical stability is a marker of low-value care. However, for commonly encountered clinical scenarios on medical units, there are no guidelines defining appropriate use criteria for laboratory tests.

OBJECTIVE

This study seeks to establish consensus-based recommendations for the utilization of common laboratory tests in medical inpatients.

DESIGN

This study uses a modified Delphi method. Participants completed two rounds of an online survey to determine appropriate testing frequencies for selected laboratory tests in commonly encountered clinical scenarios. Consensus was defined as agreement by at least 80% of participants.

PARTICIPANTS

Participants were 36 experts in internal medicine across Canada defined as internists in independent practice for ≥ 5 years with experience in medical education, quality improvement, or both. Experts represented 8 of the 10 Canadian provinces and 13 of 17 academic institutions.

MAIN MEASURES

Laboratory tests and clinical scenarios included were those that were considered common on medical units. The final survey contained a total of 45 clinical scenarios looking at the utilization of six laboratory tests (complete blood count, electrolytes, creatinine, urea, international normalized ratio, and partial thromboplastin time). The possible frequency choices were every 2-4 h, 6-8 h, twice a day, daily, every 2-3 days, weekly, or none unless there was specific diagnostic suspicion. These scenarios were reviewed by two internists with training in quality improvement and survey methods.

KEY RESULTS

Of the 45 initial clinical scenarios included, we reached consensus on 17 scenarios. We reached weak consensus on an additional 19 scenarios by combining two adjacent frequency categories.

CONCLUSIONS

A Canadian expert panel of internists has provided frequency recommendations on the utilization of six common laboratory tests in medical inpatients. These recommendations need validation in prospective studies to assess whether restrictive versus liberal laboratory test ordering impacts patient outcomes.

Comparative Impact of C-Reactive Protein Testing in Hospitalized Patients with Acute Respiratory Tract Infection: A Retrospective Cohort Study

INTRODUCTION

Evidence-based data on the usefulness of C-reactive protein (CRP) monitoring in patient outcomes are lacking. CRP testing in patients with acute respiratory tract infections (ARTIs) showed wide variability between internal medicine wards in our hospital network. In this study we aimed to investigate whether repetitive CRP tests might influence the switch of antibiotic therapy from intravenous (IV) to oral (PO) route and whether CRP measurements affect the combined outcome of readmission and in-hospital mortality.

METHODS

This was a retrospective cohort study conducted in two internal medicine wards selected in a network of five teaching hospitals on the basis of their CRP prescription frequency. Clinical and laboratory data of 296 patients with ARTIs and admitted from 1 January to 31 December 2016 were analyzed.

RESULTS

The mean ± SD of CRP tests/patient and the in-hospital length of antibiotic therapy (days) in the low-CRP (L-CRP) vs the high-CRP (H-CRP) wards were 1.14 ± 0.62 vs 3.43 ± 1.54 (p < 0.001) and 7.1 ± 2.6 vs 7.5 ± 3.2 (p = 0.298), respectively. The probability of antibiotic switching was higher in the L-CRP ward (HR 2.90, 95% CI 1.69-4.95, p < 0.001) correlating with the lower number of CRP determinations (HR 1.20, 95% CI 1.01-1.41, p = 0.034). In-hospital readmissions and mortality rates did not significantly differ between the two wards (L-CRP 17.1% vs H-CRP 10.0%, p = 0.133). The number of CRP determinations affected the combined outcome (OR 1.38, 95% CI 1.01-1.90, p = 0.043).

CONCLUSIONS

Repetitive CRP testing in ARTIs offers no added value to either antibiotic switch or patient outcomes in hospitalized patients in internal medicine wards.

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

Recurrent, inappropriate laboratory testing is a costly and wasteful use of healthcare resources. Recognizing this problem, the American Board of Internal Medicine, Canadian Society of Internal Medicine, and the Canadian Association of Pathologist all supported the Choosing Wisely campaign to reduce laboratory investigations in patients who demonstrate clinical and laboratory stability. In this narrative, we review studies looking at a variety of approaches to reduce excessive testing including education, audit and feedback, computerized physician order entry system changes, and forcing functions. Each type of intervention has its own unique advantages and disadvantages, varying in complexity, disruptiveness, effectiveness, and sustainability. Before implementing any quality improvement project, it is important to analyze the local context to identify the root causes for the practice behavior and aim to use the minimal amount of intervention to achieve the desired result. Change is often incremental and will seldom occur with a single intervention or Plan-Do-Study-Act cycle. Garnering the support of opinion leaders and a quality improvement team will help make the process and intervention a success.

Improving patient portal enrolment in an academic resident continuity clinic: quality improvement made simple

Electronic health record patient portal usage has been associated with improvement in chronic disease parameters, patient functional status and patient satisfaction. Our institution’s patient portal is a secure, online health management tool that connects patients to portions of their electronic health record.

Our quality improvement project aimed to increase patient portal enrolment significantly in our Internal Medicine resident patient panels.

This study was conducted in a large, multisite health system in Kansas City, Kansas that serves a diverse patient population. Our clinic includes 65 resident patient panels. We followed a subset of 16 resident patient panels in this quality improvement project. A baseline audit showed that 35% of the 1628 patients in these panels were enrolled in the patient portal system. A standardised, nurse-initiated portal sign-up process following patient rooming was implemented. Initial results indicated a 9.6% increase in patient portal sign-up at the end of the first 4-week cycle. We then implemented educational sessions for our clinic nurses as well as attending physicians, and achieved a 15.1% increase from baseline to the end of the second 4-week cycle, resulting in 86 patient portal activations (p<0.01).

Resident physicians worked with clinic nurse partners in two formats for this project. Nurses assigned to patient rooming for residents during the clinic sessions being studied (rooming nurses) initiated the portal sign-up process. Nurses assigned to partner with the resident for longitudinal patient care management, anchor nurses, worked with residents on items such as phone messages or portal messages. Semi-structured interviews of the four anchor nurses aligned with the 16 residents were conducted at the end of the study and revealed that nursing staff perceived increased patient portal activity to be associated with a decrease in nursing workload and an increase in patient engagement.

Optimization of Laboratory Ordering Practices for Complete Blood Count With Differential

Objectives

To investigate the utilization of CBC and CBC with differential (CBC w/diff) tests at University of Alabama at Birmingham Hospital, and to determine if a reduction in CBC w/diff tests could be achieved without negatively impacting patient care.

Methods

The quantity of testing and distribution of repeated tests before, during, and after an educational intervention were compared.

Results

CBC w/diff tests were ordered 10-fold more frequently than CBC tests. The trauma burn intensive care unit ordered the most CBC w/diff tests, with repeat tests done every 4 or 12 hours. The educational intervention reduced the number of CBC w/diff tests ordered and tests repeated every 12 hours.

Conclusions

The educational intervention changed the ordering practices of CBC w/diff and CBC tests. This was sustained after the intervention and no negative effects on patient care were noted. Similar interventions may lead to optimization of ordering practices of other laboratory tests.

Reducing the number of unnecessary routine laboratory tests through education of internal medicine residents

Introduction

The Choosing Wisely guidelines advise against ordering routine blood tests for hospitalised patients unless they change management. Unnecessary testing can lead to adverse effects (eg, iatrogenic anaemia, poor sleep quality, risk for infections and increased cost of care).

Methods

An 8-week quality initiative aimed at reducing unnecessary blood tests was implemented in three internal medicine resident inpatient services. The initiative included a 30 min educational session, reminders prior to rotation and midrotation and posters in work areas that displayed lab pricing and urged judicious testing. Residents were encouraged to justify the purpose of ordering tests in their daily progress notes. Attending physicians were made aware of the initiative. Preintervention and postintervention time points were used to compare key metrics. A >10% decrease between time periods was used as an evaluation criterion.

Results

There were 293 patient records reviewed in the preintervention period and 419 in the postintervention period. The two groups were similar in terms of age and gender. Median blood test count (complete blood count/basic metabolic profile/comprehensive metabolic profile) decreased from 4 to 2 tests per patient per day (50 % decrease) after the intervention. The median length of hospital stay decreased from 4.9 to 3.9 days (21% decrease). A decreased percentage of people requiring transfusions was also noted (2016: 6.1%, 2017: 2.9%).

Conclusion

The frequency of unnecessary routine blood tests ordered in the hospital can be decreased by educating resident physicians, making them cost conscious and aware of the indications for ordering routine labs. Frequent reminders are needed to sustain the educational benefit.

Appropriately testing patients with intermediate risk for coronary artery disease: how well are we doing?

BACKGROUND

Cardiovascular risk assessment and stratification of stable coronary artery disease represents a pivotal component of coronary artery disease management. The introduction of risk stratification tools has advanced the detection of patients with intermediate to high risk for coronary artery events. These patients will ultimately undergo non-invasive and invasive cardiovascular testing.

OBJECTIVE

To synthesise evidence illustrating risk stratification tools and non-invasive testing for patients with intermediate risk for coronary artery disease and suggest a simple method for clinicians to follow prior to ordering non-invasive cardiovascular testing.

STUDY SELECTION

Literature review of PubMed and MEDLINE Central on studies and evidence highlighting the appropriate use criteria for radionuclide myocardial perfusion imaging.

FINDINGS AND CONCLUSIONS

Overtestingusing myocardial perfusion imaging is well documented in several studies in the literature. Few studies highlighted the importance of the appropriate use criteria for radionuclide myocardial perfusion imaging, which can reduce unnecessary testing in patients with suspected coronary artery disease. Herein, we propose 'Simple Steps to Follow' to be considered by clinicians prior to ordering radionuclide myocardial perfusion imaging. The target audience of this article is internal and family medicine primary care physicians and other non-cardiologist physicians.

A Comparison of Laboratory Testing in Teaching vs Nonteaching Hospitals for 2 Common Medical Conditions

IMPORTANCE

Robust laboratory use data are lacking to support the general assumption that teaching hospitals with trainees routinely order more laboratory tests for inpatients than do nonteaching hospitals.

OBJECTIVE

To quantify differences in the use of laboratory tests between teaching and nonteaching hospitals.

DESIGN, SETTING, AND PARTICIPANTS

A cross-sectional study was performed using a statewide database to identify hospitalizations with a primary diagnosis of bacterial pneumonia or cellulitis from January 1, 2014, to June 30, 2015, at teaching and nonteaching hospitals with 100 or more hospitalizations of each condition. Patients included were adult inpatients with a primary diagnosis of bacterial pneumonia (n = 24 118) or cellulitis (n = 19 211); patients excluded were those with an intensive care unit stay, transfer from another hospital, or a length of stay that was 2 SDs or more of the condition's mean length of stay.

MAIN OUTCOMES AND MEASURES

Mean laboratory tests per day stratified by illness severity, as well as factors associated with laboratory use rates.

RESULTS

A total of 43 329 hospitalized patients (20493 women and 22836 men) had a principal diagnosis of bacterial pneumonia or cellulitis across 11 major teaching hospitals, 12 minor teaching hospitals, and 73 nonteaching hospitals in Texas. Mean number of laboratory tests per day varied significantly by hospital type and was highest for major teaching hospitals for both conditions (bacterial pneumonia: major teaching hospitals, 13.21; 95% CI, 12.91-13.51; nonteaching hospitals, 8.92; 95% CI, 8.84-9.00; P < .001; cellulitis: major teaching hospitals, 10.43; 95% CI, 10.16-10.70; nonteaching hospitals, 7.29; 95% CI, 7.22-7.36; P < .001). This association held for all levels of illness severity for both conditions, except for patients with cellulitis with the highest illness severity level. In generalized mixed linear regression models, controlling for additional patient and encounter covariates, there was a significant difference in the marginal effect of hospital teaching status on mean number of laboratory tests per day between major teaching and nonteaching hospitals (difference in marginal mean laboratory tests per day for bacterial pneumonia, 3.58; 95% CI, 2.61-4.55; P < .001; for cellulitis, 2.61; 95% CI, 1.76-3.47; P < .001).

CONCLUSIONS AND RELEVANCE

Compared with nonteaching hospitals, patients in Texas admitted to major teaching hospitals with bacterial pneumonia or cellulitis received significantly more laboratory tests after controlling for illness severity, length of stay, and patient demographics. These results support the need to examine how the culture of training environments may contribute to increased use of laboratory tests.