Health Systems Research
Oakland University William Beaumont School of Medicine
586 Pioneer Drive
Rochester, MI 48309
(248) 370-3634
Oakland University William Beaumont School of Medicine
This section includes Class of 2025 Embark Projects within the Health Systems research areas. This is a wide range of projects including Quality Improvement, Patient Safety and Health Systems studies.
Assessing Disparities in SEP-1 Bundle Compliance at Corewell Health
Zachary Awad, B.S.1, Mark Munir, B.S.1, Ramin Homayouni2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Department of Foundational Medical Studies, Oakland University William Beaumont School of Medicine, Rochester, MI
INTRODUCTION
Sepsis, a life-threatening condition resulting from a dysregulated immune response to infection, carries significant mortality. To improve sepsis management, the Centers for Medicare & Medicaid Services (CMS) introduced the SEP-1 bundle, outlining critical interventions. This study aimed to investigate disparities in SEP-1 bundle compliance, examining variations among septic patients based on sociodemographic factors like age, race, and area deprivation index, as well as clinical factors like chronic health conditions.
METHODS
This is a retrospective study of 2,235 adult sepsis patients treated at Corewell Health William Beaumont University Hospital emergency department between January 1, 2023 and December 31, 2023. Sepsis cases were identified using CMS criteria and the Quire Sepsis Algorithm. Multivariable logistic regression was used to determine the association between SEP-1 bundle compliance and patient characteristics, including age, race, gender, ADI, and 27 different pre-existing chronic conditions.
RESULTS
Using multivariable logistic regression, compliance was associated with age (aOR 1.01/year, 95% CI 1.00-1.02) and Parkinson’s disease (aOR 2.73, 95% CI 1.11-6.72), but negatively with non-ischemic heart disease (aOR 0.66, 95% CI 0.52-0.83) and rheumatoid arthritis (aOR 0.71, 95% CI 0.55-0.90). While race and gender showed no significant associations, patients in affluent areas (ADI quintiles 1 & 2) were more likely to meet SEP-1 compliance (aOR 1.39, 95% CI 1.02-1.90).
CONCLUSIONS
This study highlights the complex factors influencing SEP-1 bundle compliance in sepsis management. Age and the presence of certain chronic conditions significantly impact the likelihood of receiving timely care. Although race did not show a significant independent association in this analysis, more affluent patients were more likely to receive timely care. These findings emphasize the need for targeted interventions to improve sepsis care.
Implementation of an ERAS Program and a Standardized Closing Tray Reduces SSIs in Colorectal Operations
Elizabeth Carlson, B.S.1, Nora Laban, B.S.1, Shelli Bergeron, M.D.2, Kristin LeMarbe, M.D.2, Bradley Krasnick, M.D.2, Harry Wasvary, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Corewell Health William Beaumont University Hospital Department of Colon and Rectal Surgery , Royal Oak , MI
INTRODUCTION
Colorectal surgeries can be complicated by postoperative surgical site infections (SSIs) in up to 30% of cases. These infections increase length of stay (LOS), readmissions, reoperations, and perioperative costs. To reduce SSIs, best practice perioperative guidelines have been incorporated into practice standards. How these measures impact postoperative SSIs is the focus of this study.
METHODS
This retrospective study assesses the impact of an Enhanced Recovery After Surgery (ERAS) Program, which includes a standardized surgical bundle with variable use of a clean closing tray, on SSIs in colorectal surgeries. Inclusion and exclusion criteria for defining SSIs was determined by National Healthcare Safety Network (NHSN) criteria. Patients who had infection present at the time of surgery (PATOS) were excluded. A total of 3,451 procedures conducted between 2016 and 2021 were analyzed. A standard closing tray was used in 851 (24.7%) cases and 1,748 patients (50.7%) participated in ERAS.
RESULTS
In the ERAS group (n=1748), 68 patients had an SSI (3.9%) and 91 patients in the non-ERAS group (n=1703) had an SSI (5.3%) (p-value = 0.0417). Infection rate was decreased by 26% for ERAS participants. Use of a standardized clean closing tray (n=851) resulted in an overall SSI rate of 3.8% compared to 4.9% when a tray was not used (n=2600) (p=0.1745), a 22% SSI reduction. Adjusted odds ratio for SSI for ERAS was 0.651 (0.456-0.930) (p-value=0.018, global p-value=0.018). Adjusted odds ratio for SSI for closing trays was 0.679 (0.441-1.046) (p-value=0.079, global p-value=0.079).
CONCLUSIONS
Adherence to ERAS protocols when performing colorectal surgery significantly reduced the rate of SSIs. A reduction in SSIs was also identified when a clean closing tray was used, but this reduction was not statistically significant. The incorporation of best practice perioperative guidelines improves outcomes in colorectal surgical procedures; closing trays show promise as a measure to reduce SSIs.
Improving Treatment for Infant Perianal Abscesses
Artem Dadamyan, B.S.1, Nathan Novotny, M.D.2, Ngan Nguyen, Ph.D.3
1Oakland University William Beaumont School of Medicine, Rochester, Michigan
2Corewell Health William Beaumont University Hospital, Department of Surgery, Royal Oak, MI
3Foundational Medical Studies, Oakland University William Beaumont School of Medicine, Rochester, MI
INTRODUCTION
Perianal abscesses are tender masses located at the anal verge and develop due to infection in deep crypts of Morgagni. While surgical treatment is commonly employed, there has been increased support for non-operative management. This study aimed to compare the effectiveness of a non-operative protocol to the traditional method of incision and drainage (I&D) for management of infant perianal abscesses.
METHODS
Patients under the age of 12 months identified as having a perianal abscess less than 2cm from the anal verge and less than 3cm in size were included. Patients were treated using either the protocol or surgical approach based on surgeon preference. The protocol involves Sitz baths and warm compresses three times a day and no antibiotics. Phone call follow-ups were performed three months after initial evaluation to determine outcomes.
RESULTS
Thirteen patients under the age of one were enrolled in the study – 6 in the non-operative protocol and 7 received I&D of their abscesses. The traditional I&D group had a lower recurrence rate when compared to the non-operative protocol group 0% vs 33% (p = 0.02). No difference was seen between the groups with respect to time to resolution. All children in the non-operative protocol and traditional I&D groups had resolution of their symptoms by the three month follow up.
CONCLUSIONS
The initial findings suggest surgical management is more definitive than Sitz water baths. Surgical management showed fewer recurrences of perianal abscess after initial evaluation with a similar time to resolution as Sitz baths. While 1/3 of the non-operative protocol group needed I&D after initial surgical evaluation, 2/3 were able to be successfully managed without I&D. Although the data demonstrates that surgical management is associated with decrease in abscess recurrence in comparison to Sitz baths, both methods of treatment are effective at 3 months follow up.
Influence of Medicaid and Medicare on Total Joint Arthroplasty Outcomes: 10-year Report
Madeleine DeClercq, B.S.1, Jacob Keely, M.S.2, Robert Runner, M.D.3, Kevin M. Weisz, M.D.3, Leonardo Cavinatto, M.D.3, Drew Moore, M.D.3
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Department of Research, Oakland University William Beaumont School of Medicine, Rochester, MI
3Corewell Health William Beaumont University Hospital, Department of Orthopedic Surgery, Royal Oak, MI
INTRODUCTION
This study examines the association between insurance type—Medicaid, Medicare, and private—and short-term complications and patient-reported outcomes (PROs) following primary total joint arthroplasty.
METHODS
The Michigan Arthroplasty Registry Collaborative Quality Initiative (MARCQI) data set was queried to include patients who underwent total joint arthroplasty (TJA) from 2013-2023. Patients were categorized by insurance type and demographic information, comorbidities, 90-day outcomes, complications, and PROs were analyzed. Multivariable analyses were performed to adjust for confounders.
RESULTS
The cohort consisted of 65,300 TJA cases (49,936 patients), with 52.6% privately insured, 42.64% Medicare, and 4.77% Medicaid. Medicaid patients were younger, with higher BMIs, smoking rates, and preoperative opioid usage (p<0.001). Medicaid patients had 81.7% higher emergency department visit odds than those privately insured and 63.6% more than Medicare (p<.0001). Medicaid payer status was associated with 63.3% increased odds of developing deep vein thrombosis (DVT) compared to private insurance (p=0.0119). Medicaid and Medicare patients faced 24.3% and 31.1% greater readmission odds than privately insured (p<.0001). Medicare patients had higher odds of urinary tract infections (UTI), periprosthetic joint infections (PJI), dislocation, and fracture (p<.0001). Conversely, private-payer patients were less likely to take pre-operative medications and had fewer 90-day post-operation complications (p<.0001). Medicaid patients reported the lowest pre- and post-operative Patient Reported Outcomes Measurement Information System (PROMIS) Mental and Physical scores, Knee injury and Osteoarthritis Outcome Score (KOOS), and Hip injury and Osteoarthritis Outcome Score (HOOS), although they exhibited the greatest improvement in KOOS and HOOS scores after surgery.
CONCLUSIONS
Insurance payer type is a significantly associated with postoperative outcomes, with Medicaid and Medicare patients experiencing higher complication rates and lower PRO’s than their privately insured counterparts. These disparities underscore the necessity for tailored pre- and postoperative management in TJA patients based on insurance status.
Index Cost Comparison of Laparoscopic vs Robotic Surgery in Colon and Rectal Cancer Resection: A Retrospective Financial Investigation of Surgical Methodology Innovation at a Single Institution
Ekene Ezeokoli, M.D.1, Roula Hilli, M.D.2, Harry Wasvary, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Corewell Health William Beaumont University Hospital, Royal Oak, MI
INTRODUCTION
Robotic assisted colorectal cancer resection (R-CR) has become increasingly commonplace compared to laparoscopic cancer resection (L-CR). This study's aim was to compare the direct costs of R-CR to L-CR including relations to LOS. Patients who underwent colon and/or rectal cancer resection via R-CR or L-CR instrumentation between January 1, 2015 and December 31 2018, at our institution were evaluated. Primary outcomes were overall cost, supply cost, operating time and cost length of stay (LOS), and LOS cost. Secondary outcomes were readmission within 30 days and mortality during the surgery.
METHODS
A prospectively maintained database at William Beaumont was retrospectively queried to identify patients who underwent colon or rectal cancer resection between January 1,2015 and December 31,2018. Exclusion criteria included non-cancer resections, minimally invasive procedures converted to open, or an additional procedure from the index surgery was required during or after the initial surgery. Of the initial patients,498 patients met the final inclusion criteria. “Costs” for this study refers to the actual variable direct cost to the institution with each individual patient encounter and NOT the charges to the payor. The direct variable costs for each patient encounter included direct variable supply costs (including disposables),other direct variable costs that included operating room (OR)costs and costs related to time in the OR, and other accrued direct costs secondary to the patient’s (LOS) postoperatively.
RESULTS
240R-CR (mean age 64.9±12.4 years) and 258L-CR (mean age 66.4±15.5 years) patients met the inclusion criteria. The overall mean direct cost between R-CR and L-CR was significantly higher ($8756vs$7776 respectively, p=0.001) as well as the supply cost per case ($3789vs$2122, p < 0.001). Operating time was higher for R-CR than L-CR (224 minvs187 min, p = 0.066) but LOS was slightly lower (5.08 daysvs5.55 days, p = 0.113).
CONCLUSIONS
Cost is the main obstacle to easy and widespread use of the platform at this junction, though new developments and competition could very well reduce costs. Supply cost was the main reason for increased costs with robotic resection.
Trends in Machine Learning Algorithms for Medical Imaging: A Scoping Review
Verna Halim1, Reuben Ternes2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Oakland University, Rochester, MI
INTRODUCTION
Artificial intelligence has transformed healthcare, particularly in medical imaging, where machine learning (ML) algorithms enhance diagnostic accuracy and efficiency. However, there is limited discussion on which ML algorithms are most commonly used in medical imaging and their effectiveness across different imaging modalities. This study aims to identify the most widely used ML algorithms in medical imaging, analyze their prevalence across modalities, and evaluate reasons behind their popularity.
METHODS
A scoping review was conducted to collect and analyze peer-reviewed literature on ML algorithms in medical imaging. Relevant studies were identified through systematic searches using keywords such as "artificial intelligence," "machine learning," "deep learning," and "medical imaging" across databases. Articles were screened based on relevance to AI applications in imaging, and data was categorized according to specific algorithms used, their performance, and areas of application.
RESULTS
Neural networks, particularly deep learning models such as convolutional neural networks (CNNs), were the most commonly used algorithms in medical imaging. These models are highly effective in image classification, segmentation, and anomaly detection. Other ML techniques, including support vector machines and random forests, were also utilized, to a lesser extent. The popularity of CNNs can be attributed to their superior ability to process visual data, automate feature extraction, and achieve high diagnostic accuracy. Variability was observed across imaging modalities, with CNNs dominating radiology and histopathology, other ML approaches were more prevalent in niche applications.
CONCLUSIONS
This study highlights the dominance of deep learning techniques, particularly CNNs, in medical imaging. Their use is driven by their ability to improve diagnostic precision and efficiency. Further research is needed to evaluate biases, computational costs, and the generalizability of these algorithms across diverse patient populations and imaging modalities. These findings contribute to a broader understanding of AI trends in healthcare and may inform developments in medical imaging technology.
Beyond the Epi-pen: Prehospital Use of Epinephrine In Anaphylaxis
Randall Hilleary, B.A.1, Corey Shafer, B.S.1, Margaret Menoch2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Corewell Health William Beaumont University Hospital, Royal Oak, MI
INTRODUCTION
Prehospital epinephrine administration has been historically low by individuals having an anaphylactic reaction, their caregivers, or emergency medical service (EMS) personnel. This study aimed to determine if there was a significant difference between age, sex, or race demographics for people with allergic or anaphylactic reactions and whether there was appropriate administration of epinephrine in the prehospital setting for those arriving specifically via EMS.
METHODS
Using a retrospective chart review of all patients from three different hospitals in the Corewell Health East system from January 2015 to December 2020, descriptive statistics all patients with a coded diagnosis of anaphylaxis or allergic reaction were collected. Additionally, run reports of the patients arriving via EMS were analyzed to collect data on medication administration. Subsequently, descriptive statistics were performed.
RESULTS
A total of 20,922 patients had a coded diagnosis of anaphylaxis or allergic reaction and were treated in one of the three emergency departments that were included in the study. Of that, 1,381 of those patients arrived to the emergency department via ambulance during this time period. Our demographic analysis indicated that of the patients being treated in the emergency department, the majority of patients were primarily young Caucasian females. Additionally, 12.5% of patients required epinephrine in the emergency department having not received it in the prehospital setting. Of patients who appropriately received epinephrine in the prehospital setting, only 5% required a repeat dose in the emergency department.
CONCLUSIONS
Overall, there was good use of epinephrine in the prehospital setting. The small percentage of patients requiring repeat doses of epinephrine at the hospital be due to factors including short transport time, and improvement during transport. Importantly, the population who are receiving epinephrine at the hospital represents need for increasing education in the prehospital setting for individuals of all populations and EMS personnel.
Value Comparison of Endoscopic Carpal Tunnel Release In Two Surgical Settings
Alex Lindahl, B.S.1, Charles S. Day, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Henry Ford Health, Detroit, MI
INTRODUCTION
Prior studies have demonstrated that transitioning surgeries from a hospital outpatient
department (HOPD) to an ambulatory surgical center (ASC) lowers costs. With
500,000 carpal tunnel release (CTR) surgeries annually, CTR offers an opportunity to
determine the value of one of the most commonly performed upper extremity surgeries.
We aim to quantify the value of an endoscopic CTR in an HOPD compared to an ASC
by analyzing differences in costs and patient outcomes. We hypothesize the ASC will
provide greater value by lowering costs while maintaining patient outcomes.
METHODS
Patients were prospectively recruited from one surgeon’s practice, spanning three
clinic locations, and allocated to either cohort based on patient preference. Total costs
were comprised of time-driven activity-based labor costs (TDABC), activity-based
supply costs (ABC), and claims-based facility costs. Differences in pre-operative and
three-month post-operative PROMIS UE and PI scores were calculated to determine
PROMIS-adjusted life-years (PALYUE/PI). Total costs were divided by PALYs for each
PROM to calculate the ValueUE/PI of each cohort. The magnitude of the difference in
value between cohorts was elucidated by calculating incremental cost-effectiveness
ratios (ICERs).
RESULTS
25 Patients comprised each cohort. The ASC generated 28% lower costs compared to
the HOPD ($3,370.73 ± $128.80 vs $4,654.75 ± $140.19) (p<0.001). Average PALYUE and
PALYPI gain was not significantly greater for patients at the ASC compared to the
HOPD (PALYUE 1.06 vs 0.89; PALYPI 1.22 vs 0.92). The ASC demonstrated 40-45%
greater value, represented by a lower cost/PALYs, compared to the HOPD (ValueUE
$3,168.81/PALY vs $5,242.78/PALY, ValuePI $2,759.90/PALY vs $5,038.04/PALY).
CONCLUSIONS
We observed between 40-45% greater value by performing CTRs in the ASC. While
ASCs lowered costs by 28%, costs alone do not fully explain the value differential. This
information can guide patient-level and system-level decision-making regarding
operative treatment options for carpal tunnel syndrome.
Epic Best Practice Alerts Increases SEP1 Order Set Usage
Mark Munir, B.S.1, Zachary Awad, B.S.1, Ramin Homayouni, Ph.D2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Foundational Medical Studies, Oakland University William Beaumont School of Medicine, Rochester, MI
INTRODUCTION
Sepsis is a system-wide inflammatory response to an infection that can lead to multisystem organ failure. Sepsis has a very high mortality rate of 20-50% and is the 10th leading cause of death in the United States. Electronic medical records have implemented alert models to facilitate quicker detection of sepsis and earlier initiation of treatment for septic patients which has been shown to improve outcomes. The goal of this study is to determine if an Epic Best Practice Alert (BPA) of 5 is correlated with an increased likelihood of clinicians using the sepsis orderset.
METHODS
This project is a retrospective, observational study of 2235 patients (ages 18 and older) who had a sepsis diagnosis code and the sepsis onset was within 24 hours of arrival to the Corewell Health University Hospital emergency room between January 1 to December 31, 2023. The time sepsis onset (T0) was based on CMS Severe Sepsis criteria and identified using a proprietary Quire Sepsis Algorithm. Statistical analysis was performed using multivariable regression for determining the association between variables and outcomes. The outcomes were activation of the sepsis orderset in the EMR or compliance with SEP-1 guidelines.
RESULTS
After adjusting for age, race, gender and comorbid chronic conditions, the sepsis score of 5 is significantly (aOR 1.71, 95% CI 1.43-2.03) associated with utilization of the orderset. Other factors that were positively associated with orderset usage were age, Parkinson’s Disease and anemia.
CONCLUSIONS
Although a score threshold of 5 for the sepsis BPA produced a large number of false negatives (54.4%), it was significantly associated with the likelihood of using the sepsis order set in the EMR. Others have shown that this threshold produces many false-positives. Thus, it is important to develop models that will trigger an alert while lowering the number of false results.
A Quantitative Analysis on the Locations of Phase 3 Retina Trials in the United States over the past 5 Years
Shahrukh Naseer, B.S.1, Tarek Hassan, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Department of Ophthalmology, Corewell Health William Beaumont University Hospital
Associated Retinal Consultants
INTRODUCTION
The availability of comprehensive data on institutional participation in phase 3 clinical retina trials is limited. While the US National Library of Medicine (NLM) database catalogs individual research sites, quantifying and visualizing this information remains challenging. This study systematically analyzes research site distribution, trial participation frequency, and state-level involvement in phase 3 clinical retina trials to better understand clinical research activity.
METHODS
A retrospective analysis examined phase 3 clinical retina trials registered between December 6, 2016, and December 6, 2021. Study titles, research site names, and locations were scraped from the NLM database. The data were imported into Excel for quantitative analysis, assessing trial distribution and geographic trends.
RESULTS
Seventy-six phase 3 clinical retina trials accounted for 2,996 study sites across 44 states. Non-participating states included Alaska, Delaware, Montana, New Hampshire, North Dakota, and Wyoming. The highest trial concentrations were in California (n=65), Texas (n=65), and Florida (n=59). A total of 709 research sites participated in 3.06 trials each on average. The mean number of trials per study was 39.29. Most sites (n=380) participated in only one trial, while 16 sites were involved in 15 or more. The most frequent research sites included the Southeast Retina Center (GA, 25 trials) and the Retina Research Institute of Texas (TX, 21 trials). State population showed a moderate correlation with research site numbers (R2=0.6589) but a stronger correlation with trial numbers (R2=0.8929).
CONCLUSIONS
Findings indicate that trial volume is higher in more populous states. Increased trials within a study enhances enrollment, retention, and funding, while multi-site studies improve reproducibility and clinical translation. Furthermore, as sites and trials recruit patients from diverse racial and socioeconomic backgrounds, research power improves. These insights may inform strategies to expand access to phase 3 retina trials in underserved regions.
Door in Door out Times for Anterior Large Vessel Occlusion Stroke: A Single Multi-Site Institution’s Experience During COVID-19
Gregory Palmateer, B.A.1, Anthony Gasparotto, D.O.2, Angelique Manasseh, M.D.2, Chris Kazmierczak, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester MI
2Corewell Health, Royal Oak, MI
INTRODUCTION
Mechanical Thrombectomy (MT) is the standard of care for acute large vessel occlusion ischemic stroke; however, not all hospitals have resources to offer MT. Prompt patient transfer from primary stoke centers (PSC) to thrombectomy capable stroke centers (TSC) or comprehensive stroke centers (CSC) that offer MT is imperative. Door in Door out (DIDO) time is an established metric that reflects PSC efficiency in patient transfer. We sought to identify, compare, and analyze components of DIDO for two PSC (Hospital A and B) within the Beaumont health system transferring patients to the same CSC and find if there existed a difference.
METHODS
Utilizing our institution’s stroke registry, all patients who had a diagnosis of acute stroke and had gone a hospital transfer between December 2020 and July 2022 were identified. Patients transferred to a different institution, suffered hemorrhagic stroke, had posterior circulation stroke, or didn’t undergo MT were excluded. Data was retrospectively abstracted from patient charts. Results were analyzed via SAS v9.4.
RESULTS
Hospital A (n=22) and Hospital B (n=9) were found to have a median DIDO time of 113.5 (93.0, 136.0) and 162.0 (139.0, 215.0) and median total imaging time from CT order time to CT angiography completion of 35.0 (26.0, 49.0) and 108.0 (74.0, 122.0) respectively. Median ambulance dispatch time to patient discharge was prolonged in both groups with a time of 45.5 (40.0, 58.0) and 38.0 (29.0, 52.0) respectively.
CONCLUSIONS
There was a notable difference in DIDO times comparing Hospitals A and B. Process improvements to reduce imaging time and delays in patient transportation would likely yield the greatest effects on patient outcomes. Further research is warranted to establish a standardized target DIDO time for PSCs.
The Effect of the Adoption of the National Accreditation Program for Rectal Cancer Process on Compliance Standards at a Single Institution
Garrett W Peters, B.S.1, Gregory Thomas, M.D.2, Jacob A Applegarth, M.D.2, Forrest Bohler, B.S.1, Shelli Bergeron2, Harry J Wasvary, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Corewell Health William Beaumont University Hospital, Royal Oak, MI
INTRODUCTION
The National Accreditation Program for Rectal Cancer (NAPRC) was developed to enhance the quality of rectal cancer care in the United States. This project compared NAPRC compliance at a single tertiary care academic hospital before and after the institution adopted these standards in 2019.
METHODS
Rectal cancer patients from 2016-2023 who met NAPRC eligibility criteria were retrospectively reviewed for compliance with pre-selected patient care standards. Patients diagnosed prior to August 1, 2019 (pre-NAPRC) were compared with those diagnosed afterward (post-NAPRC) to determine whether compliance with these standards differed following the institution’s adoption of new guidelines.
RESULTS
This study included 353 patients, 146 pre-NAPRC and 207 post-NAPRC. The post-NAPRC group demonstrated significantly higher compliance with pretreatment standards compared to the pre-NAPRC group, including attaining magnetic resonance imaging (MRI) (P = 0.015), computed tomography (CT) (P < 0.001), and a carcinoembryonic antigen (CEA) level (P < 0.001). Postoperative standards were more frequently met in the post-NAPRC group regarding the photographing of surgical specimens (P < 0.001). No significant differences were observed in confirming a tissue diagnosis, starting treatment within a 60-day timeframe, or completing surgical pathology reports. Prior to initiation of the NAPRC process, the institution had achieved accreditation-level compliance in two of the seven standards. Within two years of adopting NAPRC standards, complete compliance was met in six of the seven measures.
CONCLUSIONS
A single institution’s adoption of NAPRC standards improved compliance with multiple rectal cancer care standards, achieving near-complete accreditation-level compliance within two years.
Suture Management Practices in Central Venous Catheters: A Survey Among Healthcare Professionals
Matthew Prevost, B.S.1, Nishaki Mehta, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Department of Cardiology, Corewell Health William Beaumont University Hospital, Royal Oak, MI
INTRODUCTION
Traditional suture methods for securing central venous catheters (CVCs) lack standardization, are nonadjustable post-securement, and require manual dexterity, thus heightening the risk of complications such as catheter micropistoning and central line-associated bloodstream infections (CLABSIs). This study investigates medical practitioners' experiences and challenges with CVC suturing. Our hypothesis predicts that variations in responses and differences in practitioner training may influence experiences and complication rates.
METHODS
A cross-sectional survey gathered responses from physicians at Corewell Health (Beaumont)-affiliated hospitals involved in the placement and removal of CVC sutures, alongside nurses (RNs) and nurse practitioners (NPs) involved directly in CVC removal. Chi-Square and Fisher's Exact Tests identified response variations, with questions covering topics like securement techniques, adjustment procedures, and complications.
RESULTS
Our study involved 106 participants, with 47% of physicians reporting formal training for CVC suturing. While fellows exhibited greater knowledge of multiple techniques, 51% of respondents knew only one suture securement method. Readjustment was identified as the most challenging step by 49% of participants. Fellows consistently spent more time suturing (P=0.0034) and reported more knowledge of various suturing techniques. Residents (62.1%) and attendings (58.3%) reported the highest incidence of skin nicks/cuts (at least 1 time), followed by interns (40%) and fellows (30.8%). Physicians with formal training differed significantly in suture size choice (P=0.0154). However, no significant differences were detected in readjustment frequency (P=0.6564) or skin cut incidence (P=0.7990) between formally trained and untrained physicians.
CONCLUSIONS
Our results provide firsthand medical provider experiences, revealing novel qualitative insights unexplored in existing literature. Prioritizing uniform training, safe educational interventions, and implementing simulation-based training could mitigate challenges while enriching understanding and guiding improvements in CVC suture practice. Moreso, development of innovative sutureless securement devices (SSDs) could reduce complications and enhance CVC proficiency.
Treatment Rates in a Two-Step Clostridioides difficile Diagnostic Testing Algorithm
Renee Ringler1, Matthew Sims, M.D.2,3
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Section of Infectious Disease and Internal Medicine, Department of Medicine Corewell Health William Beaumont University Hospital, Royal Oak, MI
3Department of Foundational Studies, Oakland University William Beaumont School of Medicine, Rochester, MI
INTRODUCTION
Clostridioides difficile testing has varied over time. Methods have included toxigenic culture, enzyme-linked immunoassays (EIA), and PCR. Current recommendations favor use of a two or three-step algorithm to increase the sensitivity. PCR is too sensitive, often identifying patients who are colonized with C. difficile but not experiencing active infection. To minimize the mislabeling of colonized individuals as infected, Beaumont hospitals changed to a two-step EIA algorithm. The two-step EIA tests for GDH, which indicates C. difficile carriage, and toxin which indicates C. difficile infection (CDI). Thus, a GDH+ Toxin- patient is usually colonized with C. difficile but not experiencing CDI. This study examines the treatment of patients tested for C. difficile under the new two-step algorithm.
METHODS
This study is a retrospective chart review of three Corewell hospitals. Patients aged 18+ admitted and tested for CDI from March 2021 through November 2022 were included in the study. Test results breakdown, demographics, and treatment data were collected and analyzed.
RESULTS
8355 tests were conducted during the study period from 7732 hospital encounters. Of these tests, 77% were GDH- Toxin- (negative), 15% were GDH+ Toxin- (indeterminate), 8% were GDH+ Toxin+ (positive) and 0% were GDH- Toxin+ (error). GDH+ Toxin- tests had treatment ordered in 67% of cases with the highest rates of treatment occurring in patients under 45 years of age (70.9%), while GDH- Toxin- tests had treatment ordered in 33% of cases.
CONCLUSIONS
Despite a change in testing to reduce the treatment of patients colonized with C. difficile, two-thirds of patients are still treated for CDI. Historical data shows that two-thirds of GDH+ Toxin- patients would be PCR+. Thus, despite the change in testing, there has been no appreciable reduction in over-treatment. Further education as well as additional diagnostic and antibiotic stewardship would be needed to reduce inappropriate treatment.
Implementation of an Opioid Free Pediatric Hernia Repair Protocol
Corey Shafer, B.S.1, Antonela Muca, M.D.2, Randy Hilleary, M.D.1, Kristin LeMarbe, M.D.2, Pavan Brahmamdam, M.D.3, Begum Akay, M.D.3
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Corewell Health William Beaumont University Hospital, Royal Oak, MI
3Corewell Children's at William Beaumont University Hospital
Oakland University William Beaumont School of Medicine, Rochester, MI
INTRODUCTION
We aim to decrease peri-operative opioid usage in pediatric patients undergoing routine ambulatory hernia repairs while providing similar or improved pain control through the implementation of a novel multimodal standardized pain management protocol utilizing non-narcotic analgesics, nerve blocks/caudals, and child life interventions.
METHODS
Intra-operative and post-operative opioid use was compared between patients before and after protocol implementation using equal variance two sample t-tests and Fisher Exact p-values with SAS/STAT software. High protocol adherence was assigned to cases with no intra-operative or post-operative opioid use; medium adherence was assigned to cases with no intra-operative opioid use but with post-operative opioid use, and low adherence was assigned to cases with intra-operative opioid use.
RESULTS
After seven months of protocol implementation, 175 cases were included for preliminary data analysis with 103 pre-implementation and 72 post-implementation. There were no significant differences between groups regarding patient demographics or hernia type. The median age of the population studied was 4.00 years (IQR 1.60, 6.00) and the median weight of the population studied was 17.60 kg (IQR 11.97, 24.90). Post-implementation, there was an increase in the number of nerve blocks performed (37 (51.4%) vs. 6 (5.8%), p=<.0001) and an increase in non-narcotic pain medication use intra-operatively (49 (68.1%) vs. 23 (22.3%), p=<.0001). Similarly, there was an increase in non-narcotic pain medication use post-operatively in the post-implementation group (22 (31.0%) vs. 53 (51.0%), p=.0016). There was no significant difference in post-operative complications including nausea and vomiting. Post-implementation, 63.9% of cases had either high or medium protocol adherence.
CONCLUSIONS
Initial data/protocol analysis shows a significant decrease in peri-operative opioid use and increase in use of non-narcotic medications after protocol implementation. Further interventions are needed to improve protocol adherence.
Examining the Spiritual Care Needs of Patients Preceding Death
Kevin A. Van, B.S.1, Kevin Hickey, M.S.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Department of Foundational Medical Studies, Oakland University William Beaumont School of Medicine, Rochester, MI
INTRODUCTION
Patients who die in the hospital setting present varied spiritual and emotional needs that require an array of support strategies from a spiritual care (SC) standpoint. Assessing the SC needs of dying patients within the hospital is not a well-developed area of research that lacks quantifiable data. This study primarily aims to assess patient needs and outcomes of SC prior to death following short-term (ST) and long-term (LT) hospitalizations. Similarities and differences of the needs/interventions provided through chaplaincy services among the two populations are evaluated as the secondary objective.
METHODS
Retrospective data from chaplaincy services across all units at Corewell Health William Beaumont University Hospital was obtained for the year 2023. Inclusion criteria included death within the hospital with ≥ 1 completed SC visit occurring during the same admission. Exclusion criteria included ≥ 14 days since admission without a SC visit. A total of 147 eligible subjects were subdivided into ST stay with time between admission to death ≤ 7 days (52 subjects) and LT stay with time between admission to death > 7 days (95 subjects). Predominant emotional and spiritual-based needs/outcomes were evaluated with chi-square testing to determine whether statistical similarities or differences existed.
RESULTS
There was no statistically significant difference between ST and LT stays for the predominant emotional needs (functional loss, uncertainty, and sadness) and spiritual needs (healing, hope, peace, end-of-life-issues, and dignity) as well as SC outcomes. However, there was a statistically significant difference in the emotional variables of ‘overwhelm’ (p = 0.034) and ‘fatigue’ (p = 0.042).
CONCLUSIONS
Patients who die in the hospital following a ST vs. LT stay do not exhibit significant evidence of contrasting needs related to most spiritual and emotional variables assessed; however, exploration into possible connections and interventions pertaining to ‘overwhelm’ and ‘fatigue’ in ST hospitalizations of end-of-life patients is warranted.
Consequences of Changing Frequency of Epidural Steroid Injections for Elderly Patients with Lumbar Radiculopathy
Nitin Venkatesh, B.S.1, Rebecca Clemans, M.D.2
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Corewell Health William Beaumont University Hospital, Royal Oak, MI
INTRODUCTION
Lower back pain is a significant cause of disability and can influence a patient’s quality of life (QOL) and function. A significant interventional tool are epidural steroid injections (ESIs). Previously, at Beaumont, patients with chronic lumbar radiculopathy were administered one injection every 8 weeks. However, as of December 2021, the Medicare Coverage of ESIs limited ESIs to a frequency of one injection every 12 weeks. This study aims to understand whether decreasing the frequency of ESIs from once every 8 weeks to every 12 weeks will result in changes to pain scores and impact QOL metrics such as activity levels, mood and sleep.
METHODS
This study involves administering a prospective survey to which patients self-report pain scores and impact to QOL metrics with numerical responses on a 10-point scale. A score of 10 indicates the most intense pain or pain that severely limits the mentioned activity. A score of 0 indicates absence of pain or pain that has no impact on the mentioned activity. The baseline survey is completed in clinic prior to receiving an ESI. Additional surveys are sent out every four weeks via email over the 12-week study period for a total of 4 surveys.
RESULTS
Results show that at the 8-week mark, patient’s pain scores and the impact on their QOL metrics are both ~78% of their pre-injection levels (n=3). Numerically, the total pain score at baseline is 62 versus 48 at 8 weeks. The total impact to QOL at baseline is 101 versus 79 at 8 weeks.
CONCLUSIONS
The results show that the pain levels and QOL metrics of patients receiving ESIs gets close to pre-injection levels at the 8-week mark. Although the sample size is limited, the findings suggests that patients would benefit from ESIs every 8 weeks rather than every 12.
Surgical Site Infection Bundle for Babies - Preliminary Data
Cullen Woodley, B.S.1, John Nida, M.D.2, Alexis Lonjin, M.D.1, Katelyn Ward, M.D.2, Diane Studzinski, 2, Begum Akay, M.D.3
1Oakland University William Beaumont School of Medicine, Rochester, MI
2Corewell Health William Beaumont University Hospital, Royal Oak, MI
3Corewell Health Children’s at William Beaumont University Hospital, Royal Oak, MI
INTRODUCTION
Perioperative care bundles show promise in reducing surgical site infections (SSIs) in adults but are not often used in infant surgery. Infections in babies after surgery are less common than adults but carry significant morbidity in a fragile patient. We aim to evaluate the effects of a broad perioperative care bundle in patients less than 6 months old undergoing abdominal or thoracic surgeries.
METHODS
The perioperative bundle was developed in a multi-disciplinary fashion and addressed several areas of care: standardized bowel preparation when appropriate, preoperative skin cleansing, antibiotic choice and timing, temperature control, intraoperative skin preparation, and closing protocols. Patient characteristics and outcomes were compared pre- and post-implementation of the bundle. An SSI bundle checklist was used to help confirm bundle adherence.
RESULTS
After 6 months, 44 cases were reviewed with 26 pre- and 18 post-implementation. The median estimated gestational age at birth was 36.6 weeks (IQR 30.8, 37.6). The median weight at the time of surgery was 3.3 kg (IQR 2.3, 4.0). After bundle implementation, we saw statistically significant increases in pre-operative skin cleansing (0.0% vs 61.1%, p<0.001), Chloraprep use (3.8% vs 61.1%, p<0.001), and closing protocols (0.0% vs 33.3%, p=0.0026). No difference in SSI was appreciated. The bundle was used appropriately 83% of the time.
CONCLUSIONS
A perioperative SSI bundle in babies can change the approach to infection control during surgery. The rate of SSI in both groups was low. Adoption of our bundle was quite good, but we expect its use requires more patients to appreciate its benefits.
