Faculty Bibliography

Importance: Accurate, real-time case identification is needed to target interventions to improve quality and outcomes for hospitalized patients with heart failure. Problem lists may be useful for case identification but are often inaccurate or incomplete. Machine-learning approaches may improve accuracy of identification but can be limited by complexity of implementation. Objective: To develop algorithms that use readily available clinical data to identify patients with heart failure while in the hospital. Design, Setting, and Participants: We performed a retrospective study of hospitalizations at an academic medical center. Hospitalizations for patients 18 years or older who were admitted after January 1, 2013, and discharged before February 28, 2015, were included. From a random 75% sample of hospitalizations, we developed 5 algorithms for heart failure identification using electronic health record data: (1) heart failure on problem list; (2) presence of at least 1 of 3 characteristics: heart failure on problem list, inpatient loop diuretic, or brain natriuretic peptide level of 500 pg/mL or higher; (3) logistic regression of 30 clinically relevant structured data elements; (4) machine-learning approach using unstructured notes; and (5) machine-learning approach using structured and unstructured data. Main Outcomes and Measures: Heart failure diagnosis based on discharge diagnosis and physician review of sampled medical records. Results: A total of 47119 hospitalizations were included in this study (mean [SD] age, 60.9 [18.15] years; 23 952 female [50.8%], 5258 black/African American [11.2%], and 3667 Hispanic/Latino [7.8%] patients). Of these hospitalizations, 6549 (13.9%) had a discharge diagnosis of heart failure. Inclusion of heart failure on the problem list (algorithm 1) had a sensitivity of 0.40 and a positive predictive value (PPV) of 0.96 for heart failure identification. Algorithm 2 improved sensitivity to 0.77 at the expense of a PPV of 0.64. Algorithms 3, 4, and 5 had areas under the receiver operating characteristic curves of 0.953, 0.969, and 0.974, respectively. With a PPV of 0.9, these algorithms had associated sensitivities of 0.68, 0.77, and 0.83, respectively. Conclusions and Relevance: The problem list is insufficient for real-time identification of hospitalized patients with heart failure. The high predictive accuracy of machine learning using free text demonstrates that support of such analytics in future electronic health record systems can improve cohort identification.

There is an active public debate about whether patients' socioeconomic status should be included in the readmission measures used to determine penalties in Medicare's Hospital Readmissions Reduction Program (HRRP). Using the current Centers for Medicare and Medicaid Services methodology, we compared risk-standardized readmission rates for hospitals caring for high and low proportions of patients of low socioeconomic status (as defined by their Medicaid status or neighborhood income). We then calculated risk-standardized readmission rates after additionally adjusting for patients' socioeconomic status. Our results demonstrate that hospitals caring for large proportions of patients of low socioeconomic status have readmission rates similar to those of other hospitals. Moreover, readmission rates calculated with and without adjustment for patients' socioeconomic status are highly correlated. Readmission rates of hospitals caring for patients of low socioeconomic status changed by approximately 0.1 percent with adjustment for patients' socioeconomic status, and only 3-4 percent fewer such hospitals reached the threshold for payment penalty in Medicare's HRRP. Overall, adjustment for socioeconomic status does not change hospital results in meaningful ways.

Purpose: To characterize the utility of shoulder imaging in the outpatient setting; Define predictor variables for useful shoulder imaging in terms of guiding the selection of the primary diagnosis and treatment Materials and Methods: We conducted a retrospective review of adult patients over a 32 month period evaluated and imaged for a primary complaint of shoulder pain in one of three outpatient settings: (1) orthopaedics, (2) emergency department [ED], and (3) internal medicine [IM]. Our sample population was chosen through a review of electronic medical records, using shoulder related ICD-9 codes and physician names. The main outcome variable for this study was imaging utility. A useful imaging examination was defined as a study that satisfied at least one of the following 4 criteria: changed the clinical diagnosis, guided a change in treatment selection, provided a final diagnosis, or guided definitive treatment. A utility score was assigned to each study based on the number of criteria satisfied (range 0-4) with a score of 0 defined as no utility, 1 low utility, 2 moderate utility, and score of > 3 high utility. For patients receiving multiple sequential imaging studies during their workup, each study was included and scored separately. The potential predictor variables evaluated for useful imaging included age, gender, trauma history, symptom chronicity, and injury setting (sports vs. non-sports). Statistical analysis included 95 % confidence intervals and binary logistic regression. Results: A total of 122 patients (70 female/52 male; mean age 47 years (range 18-84)) underwent a total of 171 imaging studies (109 radiographs/57 MRIs/3 CT/2 ultrasound) as part of their initial workup. 106 studies were ordered from orthopaedics, 64 from ED, and 1 from IM. CT and ultrasound utility were not assessed due to low number of cases. Overall, 95.9 % of the imaging studies met the minimum criteria for utility, most commonly helping guide the selection of a definitive treatment (71.9 %). 30.4 % of the studies were categorized as moderately useful, while 12.9 % were classified as highly useful. 95.4 % of radiographs met the criteria for utility, the majority of which were categorized as low utility (78.9 %). Both sports related injury history and trauma were predictive of at least moderate utility for radiographs (p = 0.039, p = 0.004). Younger age was a significant predictor of at least moderate utility for radiographs, most commonly in patients under 32 (p=0.003, AUC 0.748). 96.5 % of MRIs met the criteria for utility, the majority of which were categorized as moderate or high utility (84.2 %). None of the variables investigated were found to significantly predict MRI utility. For patients undergoing radiographs and MRI, MRI was found more useful than radiographs in 53 % of patients with an average utility score of 2.1. Equal utility was found in 42 % of cases, while radiographs were found most useful in 5 % of patients. Conclusion: Our study suggests that both radiographs and MRI have utility in the outpatient evaluation of shoulder pain. This serves as a potential first step towards the development of evidence based imaging algorithms that can be used and tested in future studies

Programs from the Centers for Medicare and Medicaid Services simultaneously promote strategies to lower hospital admissions and readmissions. However, there is concern that hospitals in communities that successfully reduce admissions may be penalized, as patients that are ultimately hospitalized may be sicker and at higher risk of readmission. We therefore examined the relationship between changes from 2010 to 2013 in admission rates and thirty-day readmission rates for elderly Medicare beneficiaries. We found that communities with the greatest decline in admission rates also had the greatest decline in thirty-day readmission rates, even though hospitalized patients did grow sicker as admission rates declined. The relationship between changing admission and readmission rates persisted in models that measured observed readmission rates, risk-standardized readmission rates, and the combined rate of readmission and death. Our findings suggest that communities can reduce admission rates and readmission rates in parallel, and that federal policy incentivizing reductions in both outcomes does not create contradictory incentives.

Since its publication in 2008, SQUIRE (Standards for Quality Improvement Reporting Excellence) has contributed to the completeness and transparency of reporting of quality improvement work, providing guidance to authors and reviewers of reports on healthcare improvement work. In the interim, enormous growth has occurred in understanding factors that influence the success, and failure, of healthcare improvement efforts. Progress has been particularly strong in three areas: the understanding of the theoretical basis for improvement work; the impact of contextual factors on outcomes; and the development of methodologies for studying improvement work. Consequently, there is now a need to revise the original publication guidelines. To reflect the breadth of knowledge and experience in the field, we solicited input from a wide variety of authors, editors and improvement professionals during the guideline revision process. This Explanation and Elaboration document (E&E) is a companion to the revised SQUIRE guidelines, SQUIRE 2.0. The product of collaboration by an international and interprofessional group of authors, this document provides examples from the published literature, and an explanation of how each reflects the intent of a specific item in SQUIRE. The purpose of the guidelines is to assist authors in writing clearly, precisely and completely about systematic efforts to improve the quality, safety and value of healthcare services. Authors can explore the SQUIRE statement, this E&E and related documents in detail at http://www.squire-statement.org.

Importance: Feasibility, effectiveness, and sustainability of large-scale readmission reduction efforts are uncertain. The Greater New Haven Coalition for Safe Transitions and Readmission Reductions was funded by the Center for Medicare & Medicaid Services (CMS) to reduce readmissions among all discharged Medicare fee-for-service (FFS) patients. Objective: To evaluate whether overall Medicare FFS readmissions were reduced through an intervention applied to high-risk discharge patients. Design, Setting, and Participants: This quasi-experimental evaluation took place at an urban academic medical center. Target discharge patients were older than 64 years with Medicare FFS insurance, residing in nearby zip codes, and discharged alive to home or facility and not against medical advice or to hospice; control discharge patients were older than 54 years with the same zip codes and discharge disposition but without Medicare FFS insurance if older than 64 years. High-risk target discharge patients were selectively enrolled in the program. Interventions: Personalized transitional care, including education, medication reconciliation, follow-up telephone calls, and linkage to community resources. Measurements: We measured the 30-day unplanned same-hospital readmission rates in the baseline period (May 1, 2011, through April 30, 2012) and intervention period (October 1, 2012, through May 31, 2014). Results: We enrolled 10621 (58.3%) of 18223 target discharge patients (73.9% of discharge patients screened as high risk) and included all target discharge patients in the analysis. The mean (SD) age of the target discharge patients was 79.7 (8.8) years. The adjusted readmission rate decreased from 21.5% to 19.5% in the target population and from 21.1% to 21.0% in the control population, a relative reduction of 9.3%. The number needed to treat to avoid 1 readmission was 50. In a difference-in-differences analysis using a logistic regression model, the odds of readmission in the target population decreased significantly more than that of the control population in the intervention period (odds ratio, 0.90; 95% CI, 0.83-0.99; P = .03). In a comparative interrupted time series analysis of the difference in monthly adjusted admission rates, the target population decreased an absolute -3.09 (95% CI, -6.47 to 0.29; P = .07) relative to the control population, a similar but nonsignificant effect. Conclusions and Relevance: This large-scale readmission reduction program reduced readmissions by 9.3% among the full population targeted by the CMS despite being delivered only to high-risk patients. However, it did not achieve the goal reduction set by the CMS.

OBJECTIVE: Recalling a patient to repeat a radiology examination is an adverse and, in certain cases, preventable event. Our objectives were to assess the rate of patient recalls for all imaging performed in the outpatient setting at our institution and to characterize the underlying reasons for the recalls. MATERIALS AND METHODS: We performed a retrospective review of all repeat imaging requests for an inadequate initial imaging study between January 2012 and March 2015. RESULTS: We identified 100 recall requests (mean, 2.6 requests per month), for an overall recall rate of approximately 1 in 8046 ambulatory studies and 1 in 1684 MRI studies. Nearly all recalls (98%) involved adults. A total of 95% of the recalls were for MRI studies. The most common reason for a patient recall request was an incomplete examination, making up 24% of all requests. The other causes were inadequate coverage of the area of interest (22%), protocoling errors (20%), poor imaging quality (15%), additional imaging to clarify a finding (11%), insufficient contrast visualization (7%), and incorrect patient information (1%). CONCLUSION: We found that patient recalls for imaging in the outpatient setting at our institution are not common. When recalls did occur, they were most often related to the acquisition of MR images. Improved technologist education on MRI protocoling and enhanced communication between ordering clinicians and radiologists to clarify the purpose of imaging might reduce the need for repeat ambulatory imaging.