Publications by authors named "Zhenqiu Lin"

95 Publications

Are medical record front page data suitable for risk adjustment in hospital performance measurement? Development and validation of a risk model of in-hospital mortality after acute myocardial infarction.

BMJ Open 2021 Apr 9;11(4):e045053. Epub 2021 Apr 9.

National Clinical Research Center of Cardiovascular Diseases, NHC Key Laboratory of Clinical Research for Cardiovascular Medications, State Key Laboratory of Cardiovascular Disease, Fuwai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, National Center for Cardiovascular Diseases, Beijing, China

Objectives: To develop a model of in-hospital mortality using medical record front page (MRFP) data and assess its validity in case-mix standardisation by comparison with a model developed using the complete medical record data.

Design: A nationally representative retrospective study.

Setting: Representative hospitals in China, covering 161 hospitals in modelling cohort and 156 hospitals in validation cohort.

Participants: Representative patients admitted for acute myocardial infarction. 8370 patients in modelling cohort and 9704 patients in validation cohort.

Primary Outcome Measures: In-hospital mortality, which was defined explicitly as death that occurred during hospitalisation, and the hospital-level risk standardised mortality rate (RSMR).

Results: A total of 14 variables were included in the model predicting in-hospital mortality based on MRFP data, with the area under receiver operating characteristic curve of 0.78 among modelling cohort and 0.79 among validation cohort. The median of absolute difference between the hospital RSMR predicted by hierarchical generalised linear models established based on MRFP data and complete medical record data, which was built as 'reference model', was 0.08% (10th and 90th percentiles: -1.8% and 1.6%). In the regression model comparing the RSMR between two models, the slope and intercept of the regression equation is 0.90 and 0.007 in modelling cohort, while 0.85 and 0.010 in validation cohort, which indicated that the evaluation capability from two models were very similar.

Conclusions: The models based on MRFP data showed good discrimination and calibration capability, as well as similar risk prediction effect in comparison with the model based on complete medical record data, which proved that MRFP data could be suitable for risk adjustment in hospital performance measurement.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1136/bmjopen-2020-045053DOI Listing
April 2021

Association of Angiotensin-Converting Enzyme Inhibitors and Angiotensin Receptor Blockers with the Risk of Hospitalization and Death in Hypertensive Patients with Coronavirus Disease-19.

J Am Heart Assoc 2021 Feb 24:e018086. Epub 2021 Feb 24.

Section of Cardiovascular Medicine Department of Internal Medicine Yale School of Medicine New Haven CT.

Background Despite its clinical significance, the risk of severe infection requiring hospitalization among outpatients with SARS-CoV-2 infection who receive angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) remains uncertain. Methods and Results In a propensity score-matched outpatient cohort (January - May 2020) of 2,263 Medicare Advantage and commercially insured individuals with hypertension and a positive outpatient SARS-CoV-2 test, we determined the association of ACE inhibitors and ARBs with COVID-19 hospitalization. In a concurrent inpatient cohort of 7,933 hospitalized with COVID-19, we tested their association with in-hospital mortality. The robustness of the observations was assessed in a contemporary cohort (May - August). In the outpatient study, neither ACE inhibitors (HR, 0.77, 0.53-1.13, P=0.18), nor ARBs (HR, 0.88, 0.61-1.26, P=0.48), were associated with hospitalization risk. ACE inhibitors were associated with lower hospitalization risk in the older Medicare group (HR, 0.61, 0.41-0.93, P=0.02), but not the younger commercially insured group (HR, 2.14, 0.82-5.60, P=0.12; P-interaction 0.09). Neither ACE inhibitors nor ARBs were associated with lower hospitalization risk in either population in the validation cohort. In the primary inpatient study cohort, neither ACE inhibitors (0.97, 0.81-1.16; P=0.74) nor ARBs (1.15, 0.95-1.38, P=0.15) were associated with in-hospital mortality. These observations were consistent in the validation cohort. Conclusions ACE inhibitors and ARBs were not associated with COVID-19 hospitalization or mortality. Despite early evidence for a potential association between ACE inhibitors and severe COVID-19 prevention in older individuals, the inconsistency of this observation in recent data argues against a role for prophylaxis.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1161/JAHA.120.018086DOI Listing
February 2021

Temporal relationship of computed and structured diagnoses in electronic health record data.

BMC Med Inform Decis Mak 2021 02 17;21(1):61. Epub 2021 Feb 17.

Center for Outcomes Research and Evaluation, Yale New Haven Hospital, New Haven, CT, USA.

Background: The electronic health record (EHR) holds the prospect of providing more complete and timely access to clinical information for biomedical research, quality assessments, and quality improvement compared to other data sources, such as administrative claims. In this study, we sought to assess the completeness and timeliness of structured diagnoses in the EHR compared to computed diagnoses for hypertension (HTN), hyperlipidemia (HLD), and diabetes mellitus (DM).

Methods: We determined the amount of time for a structured diagnosis to be recorded in the EHR from when an equivalent diagnosis could be computed from other structured data elements, such as vital signs and laboratory results. We used EHR data for encounters from January 1, 2012 through February 10, 2019 from an academic health system. Diagnoses for HTN, HLD, and DM were computed for patients with at least two observations above threshold separated by at least 30 days, where the thresholds were outpatient blood pressure of ≥ 140/90 mmHg, any low-density lipoprotein ≥ 130 mg/dl, or any hemoglobin A1c ≥ 6.5%, respectively. The primary measure was the length of time between the computed diagnosis and the time at which a structured diagnosis could be identified within the EHR history or problem list.

Results: We found that 39.8% of those with HTN, 21.6% with HLD, and 5.2% with DM did not receive a corresponding structured diagnosis recorded in the EHR. For those who received a structured diagnosis, a mean of 389, 198, and 166 days elapsed before the patient had the corresponding diagnosis of HTN, HLD, or DM, respectively, recorded in the EHR.

Conclusions: We found a marked temporal delay between when a diagnosis can be computed or inferred and when an equivalent structured diagnosis is recorded within the EHR. These findings demonstrate the continued need for additional study of the EHR to avoid bias when using observational data and reinforce the need for computational approaches to identify clinical phenotypes.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1186/s12911-021-01416-xDOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7890604PMC
February 2021

Administrative Claims Measure for Profiling Hospital Performance Based on 90-Day All-Cause Mortality Following Coronary Artery Bypass Graft Surgery.

Circ Cardiovasc Qual Outcomes 2021 Feb 4;14(2):e006644. Epub 2021 Feb 4.

Section of Rheumatology, Department of Internal Medicine (L.G.S.) Yale School of Medicine, New Haven, CT.

Background: Coronary artery bypass graft (CABG) surgery is a focus of bundled and alternate payment models that capture outcomes up to 90 days postsurgery. While clinical registry risk models perform well, measures encompassing mortality beyond 30 days do not currently exist. We aimed to develop a risk-adjusted hospital-level 90-day all-cause mortality measure intended for assessing hospital performance in payment models of CABG surgery using administrative data.

Methods: Building upon Centers for Medicare and Medicaid Services hospital-level 30-day all-cause CABG mortality measure specifications, we extended the mortality timeframe to 90 days after surgery and developed a new hierarchical logistic regression model to calculate hospital risk-standardized 90-day all-cause mortality rates for patients hospitalized for isolated CABG. The model was derived from Medicare claims data for a 3-year cohort between July 2014 to June 2017. The data set was randomly split into 50:50 development and validation samples. The model performance was evaluated with C statistics, overfitting indices, and calibration plot. The empirical validity of the measure result at the hospital level was evaluated against the Society of Thoracic Surgeons composite star rating.

Results: Among 137 819 CABG procedures performed in 1183 hospitals, the unadjusted mortality rate within 30 and 90 days were 3.1% and 4.7%, respectively. The final model included 27 variables. Hospital-level 90-day risk-standardized mortality rates ranged between 2.04% and 11.26%, with a median of 4.67%. C statistics in the development and validation samples were 0.766 and 0.772, respectively. We identified a strong positive correlation between 30- and 90-day risk-standardized mortality rates, with a regression slope of 1.09. Risk-standardized mortality rates also showed a stepwise trend of lower 90-day mortality with higher Society of Thoracic Surgeons composite star ratings.

Conclusions: We present a measure of hospital-level 90-day risk-standardized mortality rates following isolated CABG. This measure complements Centers for Medicare and Medicaid Services' existing 30-day CABG mortality measure by providing greater insight into the postacute recovery period. It offers a balancing measure to ensure efforts to reduce costs associated with CABG recovery and rehabilitation do not result in unintended consequences.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1161/CIRCOUTCOMES.120.006644DOI Listing
February 2021

Human disease clinical treatment network for the elderly: The analysis of medicare inpatient length of stay data.

Stat Med 2021 Apr 2;40(8):2083-2099. Epub 2021 Feb 2.

Department of Biostatistics, Yale University, New Haven, Connecticut, USA.

Disease clinical treatment measures, such as inpatient length of stay (LOS), have been examined for most if not all diseases. Such analysis has important implications for the management and planning of health care, financial, and human resources. In addition, clinical treatment measures can also informatively reflect intrinsic disease properties such as severity. The existing studies mostly focus on either a single disease (or a few pre-selected and closely related diseases) or all diseases combined. In this study, we take a new and innovative perspective, examine the interconnections in length of stay (LOS) among diseases, and construct the very first disease clinical treatment network on LOS. To accommodate uniquely challenging data distributions, a new conditional network construction approach is developed. Based on the constructed network, the analysis of important network properties is conducted. The Medicare data on 100 000 randomly selected subjects for the period of January 2008 to December 2018 is analyzed. The network structure and key properties are found to have sensible biomedical interpretations. Being the very first of its kind, this study can be informative to disease clinical management, advance our understanding of disease interconnections, and foster complex network analysis.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/sim.8893DOI Listing
April 2021

Analysis of Hospital Resource Availability and COVID-19 Mortality Across the United States.

J Hosp Med 2021 04;16(4):211-214

Department of Emergency Medicine, Yale School of Medicine, New Haven, Connecticut.

Although the impact of COVID-19 has varied greatly across the United States, there has been little assessment of hospital resources and mortality. We examine hospital resources and death counts among hospital referral regions from March 1 to July 26, 2020. This was an analysis of American Hospital Association data with COVID-19 data from the New York Times. Hospital-based resource availabilities were characterized per COVID-19 case. Death count was defined by monthly confirmed COVID-19 deaths. Geographic areas with fewer intensive care unit beds (incident rate ratio [IRR], 0.194; 95% CI, 0.076-0.491), nurses (IRR, 0.927; 95% CI, 0.888-0.967), and general medicine/surgical beds (IRR, 0.800; 95% CI, 0.696-0.920) per COVID-19 case were statistically significantly associated with an increased incidence rate of death in April 2020. This underscores the potential impact of innovative hospital capacity protocols and care models to create resource flexibility to limit system overload early in a pandemic.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.12788/jhm.3539DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8025594PMC
April 2021

Analysis of Hospital Resource Availability and COVID-19 Mortality Across the United States.

J Hosp Med 2021 04;16(4):211-214

Department of Emergency Medicine, Yale School of Medicine, New Haven, Connecticut.

Although the impact of COVID-19 has varied greatly across the United States, there has been little assessment of hospital resources and mortality. We examine hospital resources and death counts among hospital referral regions from March 1 to July 26, 2020. This was an analysis of American Hospital Association data with COVID-19 data from the New York Times. Hospital-based resource availabilities were characterized per COVID-19 case. Death count was defined by monthly confirmed COVID-19 deaths. Geographic areas with fewer intensive care unit beds (incident rate ratio [IRR], 0.194; 95% CI, 0.076-0.491), nurses (IRR, 0.927; 95% CI, 0.888-0.967), and general medicine/surgical beds (IRR, 0.800; 95% CI, 0.696-0.920) per COVID-19 case were statistically significantly associated with an increased incidence rate of death in April 2020. This underscores the potential impact of innovative hospital capacity protocols and care models to create resource flexibility to limit system overload early in a pandemic.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.12788/jhm.3539DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8025594PMC
April 2021

Variation in Risk-standardized Rates and Causes of Unplanned Hospital Visits Within 7 Days of Hospital Outpatient Surgery.

Ann Surg 2020 Nov 17. Epub 2020 Nov 17.

Center for Outcomes Research and Evaluation, Yale New Haven Hospital, New Haven, Connecticut.

Objectives: The objectives of this study were to compare risk-standardized hospital visit ratios of the predicted to expected number of unplanned hospital visits within 7 days of same-day surgeries performed at US hospital outpatient departments (HOPDs) and to describe the causes of hospital visits.

Summary Of Background Data: More than half of procedures in the US are performed in outpatient settings, yet little is known about facility-level variation in short-term safety outcomes.

Methods: The study cohort included 1,135,441 outpatient surgeries performed at 4058 hospitals between October 1, 2015 and September 30, 2016 among Medicare Fee-for-Service beneficiaries aged ≥65 years. Hospital-level, risk-standardized measure scores of unplanned hospital visits (emergency department visits, observation stays, and unplanned inpatient admissions) within 7 days of hospital outpatient surgery were calculated using hierarchical logistic regression modeling that adjusted for age, clinical comorbidities, and surgical procedural complexity.

Results: Overall, 7.8% of hospital outpatient surgeries were followed by an unplanned hospital visit within 7 days. Many of the leading reasons for unplanned visits were for potentially preventable conditions, such as urinary retention, infection, and pain. We found considerable variation in the risk-standardized ratio score across hospitals. The 203 best-performing HOPDs, at or below the 5th percentile, had at least 22% fewer unplanned hospital visits than expected, whereas the 202 worst-performing HOPDs, at or above the 95th percentile, had at least 29% more post-surgical visits than expected, given their case and surgical procedure mix.

Conclusions: Many patients experience an unplanned hospital visit within 7 days of hospital outpatient surgery, often for potentially preventable reasons. The observed variation in performance across hospitals suggests opportunities for quality improvement.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1097/SLA.0000000000004627DOI Listing
November 2020

Unscheduled Care Access in the United States-A Tale of Two Emergency Departments.

Am J Emerg Med 2020 Oct 27. Epub 2020 Oct 27.

Center for Outcomes Research & Evaluation, Yale University, New Haven, CT, United States of America; Section of Cardiovascular Medicine, Yale University School of Medicine, New Haven, CT, United States of America. Electronic address:

Background: Rural communities face challenges in accessing healthcare services due to physician shortages and limited unscheduled care capabilities in office settings. As a result, rural hospital-based Emergency Departments (ED) may disproportionately provide acute, unscheduled care needs. We sought to examine differences in ED utilization and the relative role of the ED in providing access to unscheduled care between rural and urban communities.

Methods: Using a 20% sample of the 2012 Medicare Chronic Condition Warehouse, we studied the overall ED visit rate and the unscheduled care rate by geography using the Dartmouth Atlas' hospital referral regions (HRR). We calculated HRR urbanicity as the proportion of beneficiaries residing in an urban zip code within each HRR. We report descriptive statistics and utilize K-means clustering based on the ED visit rates and unscheduled care rates.

Results: We found rural ED use is more common and disproportionately the site of unscheduled care delivery when compared to urban communities. The ED visit and. unscheduled care proportions were negatively correlated with increased urbanicity (r =. -0.48, p < 0.001; r = -0.58, p < 0.001).

Conclusion: The use and role of EDs by Medicare beneficiaries appears to be substantially different between urban and rural areas. This suggests that the ED may play a distinct role within the healthcare delivery system of rural communities that face disproportionate barriers to care access.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.ajem.2020.08.095DOI Listing
October 2020

The Promise of Big Data and Digital Solutions in Building a Cardiovascular Learning System: Opportunities and Barriers.

Methodist Debakey Cardiovasc J 2020 Jul-Sep;16(3):212-219

YALE SCHOOL OF MEDICINE, NEW HAVEN, CONNECTICUT.

The learning health system is a conceptual model for continuous learning and knowledge generation rooted in the daily practice of medicine. While companies such as Google and Amazon use dynamic learning systems that learn iteratively through every customer interaction, this efficiency has not materialized on a comparable scale in health systems. An ideal learning health system would learn from every patient interaction to benefit the care for the next patient. Notable advances include the greater use of data generated in the course of clinical care, Common Data Models, and advanced analytics. However, many remaining barriers limit the most effective use of large and growing health care data assets. In this review, we explore the accomplishments, opportunities, and barriers to realizing the learning health system.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.14797/mdcj-16-3-212DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7587314PMC
November 2020

Seroprevalence of SARS-CoV-2-Specific IgG Antibodies Among Adults Living in Connecticut: Post-Infection Prevalence (PIP) Study.

Am J Med 2020 Oct 29. Epub 2020 Oct 29.

Center for Outcomes Research and Evaluation, Yale New Haven Hospital, New Haven, Conn; Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, Conn; Department of Health Policy and Management, Yale School of Public Health, New Haven, Conn. Electronic address:

Background: A seroprevalence study can estimate the percentage of people with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibodies in the general population; however, most existing reports have used a convenience sample, which may bias their estimates.

Methods: We sought a representative sample of Connecticut residents, ages ≥18 years and residing in noncongregate settings, who completed a survey between June 4 and June 23, 2020, and underwent serology testing for SARS-CoV-2-specific immunoglobulin G (IgG) antibodies between June 10 and July 29, 2020. We also oversampled non-Hispanic black and Hispanic subpopulations. We estimated the seroprevalence of SARS-CoV-2-specific immunoglobulin G antibodies and the prevalence of symptomatic illness and self-reported adherence to risk-mitigation behaviors among this population.

Results: Of the 567 respondents (mean age 50 [± 17] years; 53% women; 75% non-Hispanic white individuals) included at the state level, 23 respondents tested positive for SARS-CoV-2-specific antibodies, resulting in weighted seroprevalence of 4.0 (90% confidence interval [CI] 2.0-6.0). The weighted seroprevalence for the oversampled non-Hispanic black and Hispanic populations was 6.4% (90% CI 0.9-11.9) and 19.9% (90% CI 13.2-26.6), respectively. The majority of respondents at the state level reported following risk-mitigation behaviors: 73% avoided public places, 75% avoided gatherings of families or friends, and 97% wore a facemask, at least part of the time.

Conclusions: These estimates indicate that the vast majority of people in Connecticut lack antibodies against SARS-CoV-2, and there is variation by race and ethnicity. There is a need for continued adherence to risk-mitigation behaviors among Connecticut residents to prevent resurgence of COVID-19 in this region.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.amjmed.2020.09.024DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7598362PMC
October 2020

Quality Measure Public Reporting Is Associated with Improved Outcomes Following Hip and Knee Replacement.

J Bone Joint Surg Am 2020 Oct;102(20):1799-1806

Yale-New Haven Health System Center for Outcome Research and Evaluation, New Haven, Connecticut.

Background: Given the inclusion of orthopaedic quality measures in the Centers for Medicare & Medicaid Services national hospital payment programs, the present study sought to assess whether the public reporting of total hip arthroplasty (THA) and total knee arthroplasty (TKA) risk-standardized readmission rates (RSRRs) and complication rates (RSCRs) was temporally associated with a decrease in the rates of these outcomes among Medicare beneficiaries.

Methods: Annual trends in national observed and hospital-level RSRRs and RSCRs were evaluated for patients who underwent hospital-based inpatient hip and/or knee replacement procedures from fiscal year 2010 to fiscal year 2016. Hospital-level rates were calculated with use of the same measures and methodology that were utilized in public reporting. Annual trends in the distribution of hospital-level outcomes were then examined with use of density plots.

Results: Complication and readmission rates and variation declined steadily from fiscal year 2010 to fiscal year 2016. Reductions of 33% and 25% were noted in hospital-level RSCRs and RSRRs, respectively. The interquartile range decreased by 18% (relative reduction) for RSCRs and by 34% (relative reduction) for RSRRs. The frequency of risk variables in the complication and readmission models did not systematically change over time, suggesting no evidence of widespread bias or up-coding.

Conclusions: This study showed that hospital-level complication and readmission rates following THA and TKA and the variation in hospital-level performance declined during a period coinciding with the start of public reporting and financial incentives associated with measurement. The consistently decreasing trend in rates of and variation in outcomes suggests steady improvements and greater consistency among hospitals in clinical outcomes for THA and TKA patients in the 2016 fiscal year compared with the 2010 fiscal year. The interactions between public reporting, payment, and hospital coding practices are complex and require further study.

Level Of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.2106/JBJS.19.00964DOI Listing
October 2020

Where Skilled Nursing Facility Residents Get Acute Care: Is the Emergency Department the Medical Home?

J Appl Gerontol 2020 Aug 25:733464820950125. Epub 2020 Aug 25.

Yale School of Medicine, New Haven, CT, USA.

Objectives: This study aimed to characterize the distribution of acute care visits among Medicare beneficiaries receiving skilled nursing facility (SNF) services.

Methods: We conducted a cross-sectional analysis of a 20% sample of continuously enrolled Medicare beneficiaries in the 2012 Chronic Condition Warehouse data set. Beneficiaries were grouped by the number of days of SNF services, and acute care visits were categorized as "before SNF," "during SNF," or "after SNF."

Results: Among the 10,717,786 Medicare beneficiaries analyzed, 384,312 (3.6%) had at least one SNF stay.

Discussion: Beneficiaries who received SNF services had a higher proportion of acute care visits made to emergency departments (EDs) than beneficiaries who did not receive SNF services. Also, a higher proportion of acute care visits were made to EDs by beneficiaries after a SNF stay in comparison to residents actively residing in a SNF. The acute care capabilities of SNFs and post-SNF transitions of care to the community setting are discussed.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1177/0733464820950125DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7904961PMC
August 2020

Comparison of Estimated Excess Deaths in New York City During the COVID-19 and 1918 Influenza Pandemics.

JAMA Netw Open 2020 08 3;3(8):e2017527. Epub 2020 Aug 3.

Division of Infectious Diseases, Department of Medicine, Emory University School of Medicine, Atlanta, Georgia.

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1001/jamanetworkopen.2020.17527DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7426746PMC
August 2020

Timely estimation of National Admission, readmission, and observation-stay rates in medicare patients with acute myocardial infarction, heart failure, or pneumonia using near real-time claims data.

BMC Health Serv Res 2020 Aug 10;20(1):733. Epub 2020 Aug 10.

Center for Outcomes Research and Evaluation, Yale-New Haven Health System, New Haven, CT, USA.

Background: To estimate, prior to finalization of claims, the national monthly numbers of admissions and rates of 30-day readmissions and post-discharge observation-stays for Medicare fee-for-service beneficiaries hospitalized with acute myocardial infarction (AMI), heart failure (HF), or pneumonia.

Methods: The centers for Medicare & Medicaid Services (CMS) Integrated Data Repository, including the Medicare beneficiary enrollment database, was accessed in June 2015, February 2017, and February 2018. We evaluated patterns of delay in Medicare claims accrual, and used incomplete, non-final claims data to develop and validate models for real-time estimation of admissions, readmissions, and observation stays.

Results: These real-time reporting models accurately estimate, within 2 months from admission, the monthly numbers of admissions, 30-day readmission and observation-stay rates for patients with AMI, HF, or pneumonia.

Conclusions: This work will allow CMS to track the impact of policy decisions in real time and enable hospitals to better monitor their performance nationally.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1186/s12913-020-05611-wDOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7416804PMC
August 2020

Evaluation of the Anticipated Burden of COVID-19 on Hospital-Based Healthcare Services Across the United States.

medRxiv 2020 Apr 3. Epub 2020 Apr 3.

Background: Coronavirus disease-19 (COVID-19) is a global pandemic, with the potential to infect nearly 60% of the population. The anticipated spread of the virus requires an urgent appraisal of the capacity of US healthcare services and the identification of states most vulnerable to exceeding their capacity Methods: In the American Hospital Association survey for 2018, a database of US community hospitals, we identified total inpatient beds, adult intensive care unit (ICU) beds, and airborne isolation rooms across all hospitals in each state of continental US. The burden of COVID-19 hospitalizations was estimated based on a median hospitalization duration of 12 days and was evaluated for a 30-day reporting period.

Results: At 5155 US community hospitals across 48 states in the contiguous US and Washington DC, there were a total of 788,032 inpatient beds, 68,280 adult ICU beds, and 44,222 isolation rooms. The median daily bed occupancy was 62.8% (IQR 58.1%, 66.6%) across states. Nationally, for every 10,000 individuals, there are 24.2 inpatient beds, 2.8 adult ICU beds, and 1.4 isolation beds. There is a 3-fold variation in the number of inpatient beds available across the US, ranging from 16.4 per 10,000 in Oregon to 47 per 10,000 in South Dakota. There was also a similar 3-fold variation in available or non-occupied beds, ranging from 4.7 per 10,000 in Connecticut through 18.3 per 10,000 in North Dakota. The availability of ICU beds is low nationally, ranging from 1.4 per 10,000 in Nevada to 4.7 per 10000 in Washington DC. Hospitalizations for COVID-19 in a median 0.2% (IQR 0.2 %, 0.3%) of state population, or 1.4% of state's older adults (1.0%, 1.9%) will require all non-occupied beds. Further, a median 0.6% (0.5%, 0.8%) of state population, or 3.9% (3.1%, 4.6%) of older individuals would require 100% of inpatient beds.

Conclusion: The COVID-19 pandemic is likely to overwhelm the limited number of inpatient and ICU beds for the US population. Hospitals in half of US states would exceed capacity if less than 0.2% of the state population requires hospitalization in any given month.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1101/2020.04.01.20050492DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7276011PMC
April 2020

Association of Angiotensin-Converting Enzyme Inhibitors and Angiotensin Receptor Blockers with the Risk of Hospitalization and Death in Hypertensive Patients with Coronavirus Disease-19.

medRxiv 2020 May 19. Epub 2020 May 19.

Center for Outcomes Research and Evaluation, Yale-New Haven Hospital, New Haven, CT.

Background: Whether angiotensin-converting enzyme (ACE) Inhibitors and angiotensin receptor blockers (ARBs) mitigate or exacerbate SARS-CoV-2 infection remains uncertain. In a national study, we evaluated the association of ACE inhibitors and ARB with coronavirus disease-19 (COVID-19) hospitalization and mortality among individuals with hypertension.

Methods: Among Medicare Advantage and commercially insured individuals, we identified 2,263 people with hypertension, receiving ≥1 antihypertensive agents, and who had a positive outpatient SARS-CoV-2 test (outpatient cohort). In a propensity score-matched analysis, we determined the association of ACE inhibitors and ARBs with the risk of hospitalization for COVID-19. In a second study of 7,933 individuals with hypertension who were hospitalized with COVID-19 (inpatient cohort), we tested the association of these medications with in-hospital mortality. We stratified all our assessments by insurance groups.

Results: Among individuals in the outpatient and inpatient cohorts, 31.9% and 29.8%, respectively, used ACE inhibitors and 32.3% and 28.1% used ARBs. In the outpatient study, over a median 30.0 (19.0 - 40.0) days after testing positive, 12.7% were hospitalized for COVID-19. In propensity score-matched analyses, neither ACE inhibitors (HR, 0.77 [0.53, 1.13], P = 0.18), nor ARBs (HR, 0.88 [0.61, 1.26], P = 0.48), were significantly associated with risk of hospitalization. In analyses stratified by insurance group, ACE inhibitors, but not ARBs, were associated with a significant lower risk of hospitalization in the Medicare group (HR, 0.61 [0.41, 0.93], P = 0.02), but not the commercially insured group (HR: 2.14 [0.82, 5.60], P = 0.12; P-interaction 0.09). In the inpatient study, 14.2% died, 59.5% survived to discharge, and 26.3% had an ongoing hospitalization. In propensity score-matched analyses, neither use of ACE inhibitor (0.97 [0.81, 1.16]; P = 0.74) nor ARB (1.15 [0.95, 1.38]; P = 0.15) was associated with risk of in-hospital mortality, in total or in the stratified analyses.

Conclusions: The use of ACE inhibitors and ARBs was not associated with the risk of hospitalization or mortality among those infected with SARS-CoV-2. However, there was a nearly 40% lower risk of hospitalization with the use of ACE inhibitors in the Medicare population. This finding merits a clinical trial to evaluate the potential role of ACE inhibitors in reducing the risk of hospitalization among older individuals, who are at an elevated risk of adverse outcomes with the infection.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1101/2020.05.17.20104943DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7273249PMC
May 2020

Quality of informed consent documents among US. hospitals: a cross-sectional study.

BMJ Open 2020 05 19;10(5):e033299. Epub 2020 May 19.

Center for Outcomes Research and Evaluation, Yale New Haven Health System, New Haven, Connecticut, USA.

Objective: To determine whether informed consent for surgical procedures performed in US hospitals meet a minimum standard of quality, we developed and tested a quality measure of informed consent documents.

Design: Retrospective observational study of informed consent documents.

Setting: 25 US hospitals, diverse in size and geographical region.

Cohort: Among Medicare fee-for-service patients undergoing elective procedures in participating hospitals, we assessed the informed consent documents associated with these procedures. We aimed to review 100 qualifying procedures per hospital; the selected sample was representative of the procedure types performed at each hospital.

Primary Outcome: The outcome was hospital quality of informed consent documents, assessed by two independent raters using an eight-item instrument previously developed for this measure and scored on a scale of 0-20, with 20 representing the highest quality. The outcome was reported as the mean hospital document score and the proportion of documents meeting a quality threshold of 10. Reliability of the hospital score was determined based on subsets of randomly selected documents; face validity was assessed using stakeholder feedback.

Results: Among 2480 informed consent documents from 25 hospitals, mean hospital scores ranged from 0.6 (95% CI 0.3 to 0.9) to 10.8 (95% CI 10.0 to 11.6). Most hospitals had at least one document score at least 10 out of 20 points, but only two hospitals had >50% of their documents score above a 10-point threshold. The Spearman correlation of the measures score was 0.92. Stakeholders reported that the measure was important, though some felt it did not go far enough to assess informed consent quality.

Conclusion: All hospitals performed poorly on a measure of informed consent document quality, though there was some variation across hospitals. Measuring the quality of hospital's informed consent documents can serve as a first step in driving attention to gaps in quality.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1136/bmjopen-2019-033299DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7247389PMC
May 2020

An instrument for assessing the quality of informed consent documents for elective procedures: development and testing.

BMJ Open 2020 05 19;10(5):e033297. Epub 2020 May 19.

Section of Cardiovascular Medicine, Yale School of Medicine, New Haven, Connecticut, USA.

Objective: To develop a nationally applicable tool for assessing the quality of informed consent documents for elective procedures.

Design: Mixed qualitative-quantitative approach.

Setting: Convened seven meetings with stakeholders to obtain input and feedback on the tool.

Participants: Team of physician investigators, measure development experts, and a working group of nine patients and patient advocates (caregivers, advocates for vulnerable populations and patient safety experts) from different regions of the country.

Interventions: With stakeholder input, we identified elements of high-quality informed consent documents, aggregated into three domains: content, presentation and timing. Based on this comprehensive taxonomy of key elements, we convened the working group to offer input on the development of an abstraction tool to assess the quality of informed consent documents in three phases: (1) selecting the highest-priority elements to be operationalised as items in the tool; (2) iteratively refining and testing the tool using a sample of qualifying informed consent documents from eight hospitals; and (3) developing a scoring approach for the tool. Finally, we tested the reliability of the tool in a subsample of 250 informed consent documents from 25 additional hospitals.

Outcomes: Abstraction tool to evaluate the quality of informed consent documents.

Results: We identified 53 elements of informed consent quality; of these, 15 were selected as highest priority for inclusion in the abstraction tool and 8 were feasible to measure. After seven cycles of iterative development and testing of survey items, and development and refinement of a training manual, two trained raters achieved high item-level agreement, ranging from 92% to 100%.

Conclusions: We identified key quality elements of an informed consent document and operationalised the highest-priority elements to define a minimum standard for informed consent documents. This tool is a starting point that can enable hospitals and other providers to evaluate and improve the quality of informed consent.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1136/bmjopen-2019-033297DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7247404PMC
May 2020

Availability of Telemedicine Services Across Hospitals in the United States in 2018: A Cross-sectional Study.

Ann Intern Med 2020 09 30;173(6):503-505. Epub 2020 Apr 30.

Center for Outcomes Research and Evaluation, Yale-New Haven Hospital, Yale School of Medicine, and Yale School of Public Health, New Haven, Connecticut (H.M.K.).

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.7326/M20-1201DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7212823PMC
September 2020

Cross-sectional Analysis of Emergency Department and Acute Care Utilization Among Medicare Beneficiaries.

Acad Emerg Med 2020 07 20;27(7):570-579. Epub 2020 May 20.

Yale-New Haven Hospital Center for Outcomes Research and Evaluation, New Haven, CT.

Background: We sought to develop a claims-based definition of unscheduled care to describe the use and role of the emergency department (ED) in providing unscheduled care to vulnerable older adult populations.

Methods: This study was a cross-sectional analysis of national 20% sample of Medicare beneficiaries included in the 2012 Chronic Condition Warehouse data set. We measured three outcomes: the number of ED visits per 1,000 Medicare beneficiaries, the proportion of all unscheduled ED and office-based visits occurring in the ED and the number of ED and non-ED unscheduled visits adjusting for risk factors. Each outcome was estimated for vulnerable subpopulations of Medicare beneficiaries with multiple chronic conditions (MCCs), dual eligibility, hospice enrollment, and skilled nursing facility use.

Results: A total of 10,717,786 Medicare beneficiaries were included with 33,696,461 potentially unscheduled care visits of which 5,192,235 (15%) occurred in the ED, 364,334 (1.1%) in facility-based urgent care, and 31,570,113 (84%) in ambulatory office settings. In regression analyses each subpopulation was more likely to visit the ED for unscheduled care services than the reference population of Medicare beneficiaries ages 65 to 80. Dual-eligible beneficiaries demonstrated higher ED visit rates and lower non-ED visit rates for unscheduled care. The subpopulation with MCCs uses both the ED and the non-ED setting for unscheduled care more so than any other group.

Conclusions: Medicare beneficiaries, particularly vulnerable subpopulations, disproportionately visit the ED in comparison to physician offices for unscheduled care. Efforts to improve care coordination, measure quality, or reform payment to influence ED visitation should acknowledge these patterns and the unique availability of acute care services in the ED.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1111/acem.13971DOI Listing
July 2020

Readmission and Mortality After Hospitalization for Myocardial Infarction and Heart Failure.

J Am Coll Cardiol 2020 02;75(7):736-746

Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, Connecticut; Center for Outcomes Research and Evaluation, Yale-New Haven Hospital, New Haven, Connecticut; Department of Health Policy and Management, Yale School of Public Health, New Haven, Connecticut.

Background: Readmission rates after acute myocardial infarction (AMI) and heart failure (HF) hospitalizations have decreased in the United States since the implementation of the Hospital Readmissions Reduction Program.

Objectives: This study was designed to examine the temporal trends of readmission and mortality after AMI and HF in Ontario, Canada, where reducing hospital readmissions has not had a policy incentive.

Methods: The cohort was comprised of AMI or HF patients 65 years of age or older who had been hospitalized from 2006 to 2017. Primary outcomes were 30-day readmission and post-discharge mortality. Secondary outcomes included in-hospital mortality, 30-day mortality from admission, and in-hospital mortality or 30-day mortality post-discharge. Adjusted monthly trends for each outcome were examined over the study period.

Results: Our cohorts included 152,808 AMI and 223,283 HF patients. Age- and sex-standardized AMI hospitalization rates in Ontario declined 32% from 2006 to 2017 while HF hospitalization rates declined slightly (9.1%). For AMI, risk-adjusted 30-day readmission rates declined from 17.4% in 2006 to 14.7% in 2017. All AMI risk-adjusted mortality rates also declined from 2006 to 2017 with 30-day post-discharge mortality from 5.1% to 4.4%. For HF, overall risk-adjusted 30-day readmission was largely unchanged from 2006 to 2014 at 21.9%, followed by a decline to 20.8% in 2017. Risk-adjusted 30-day post-discharge mortality declined from 7.1% in 2006 to 6.6% in 2017.

Conclusions: The patterns of outcomes in Ontario are consistent with the United States for AMI, but diverge for HF. For AMI and HF, admissions, readmissions, and mortality rates declined over this period. The reasons for the country-specific patterns for HF need further exploration.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jacc.2019.12.026DOI Listing
February 2020

EPIDEMIOLOGIC CHARACTERISTICS AND RISK FACTORS FOR CONGENITAL HYPOTHYROIDISM FROM 2009 TO 2018 IN XIAMEN, CHINA.

Endocr Pract 2020 Jun 22;26(6):585-594. Epub 2020 Jan 22.

Early diagnosis and treatment of children with congenital hypothyroidism (CH) through newborn screening can effectively prevent delayed development. This study was designed to investigate the pathogenesis and factors that influence CH in urban areas of China between 2009 and 2018. A retrospective analysis of newborn screening data and diagnosis and treatment information for CH diagnosed in the information database of the neonatal disease screening center in one of China's five special economic zones from 2009 to 2018. Of the 947,258 newborns screened between 2009 and 2018, 829 (406 girls) were diagnosed with CH at birth (1 diagnosis/1,136 births). Among the 608 cases of CH diagnosed at birth and re-evaluated at the age of 3 years, 487 were permanent congenital hypothyroidism (PCH, 1/1,429), and 121 were transient congenital hypothyroidism (TCH, 1/5,882). A total of 83.2% of infants with PCH (405/487) underwent thyroid imaging in the neonatal period, of which thyroid dysgenesis accounted for 28.64% (116/405) and functional defects accounted for 71.36% (289/405). The incidence of CH changed significantly in infants with initial serum thyroid-stimulating hormone concentrations of 41 to 100 mIU/L and ≥100 mIU/L, whereas the incidence of mild CH showed a slight increase. The incidence of CH was significantly higher in postterm infants (1/63) and low-birth-weight infants (1/370). In the past decade, the incidence of CH has increased, mainly due to the increase in the incidence of PCH and TCH. The incidence of mild CH has increased slightly. Postterm birth and low birth weight are important factors affecting the incidence of CH. = congenital hypothyroidism; = free thyroxine; = levothyroxine sodium; = permanent congenital hypothyroidism; = transient congenital hypothyroidism; = thyroid-stimulating hormone; = total thyroxine.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.4158/EP-2019-0491DOI Listing
June 2020

Post-discharge acute care and outcomes following readmission reduction initiatives: national retrospective cohort study of Medicare beneficiaries in the United States.

BMJ 2020 01 15;368:l6831. Epub 2020 Jan 15.

Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, USA.

Objectives: To determine whether patients discharged after hospital admissions for conditions covered by national readmission programs who received care in emergency departments or observation units but were not readmitted within 30 days had an increased risk of death and to evaluate temporal trends in post-discharge acute care utilization in inpatient units, emergency departments, and observation units for these patients.

Design: Retrospective cohort study.

Setting: Medicare claims data for 2008-16 in the United States.

Participants: Patients aged 65 or older admitted to hospital with heart failure, acute myocardial infarction, or pneumonia-conditions included in the US Hospital Readmissions Reduction Program.

Main Outcome Measures: Post-discharge 30 day mortality according to patients' 30 day acute care utilization; acute care utilization in inpatient and observation units and the emergency department during the 30 day and 31-90 day post-discharge period.

Results: 3 772 924 hospital admissions for heart failure, 1 570 113 for acute myocardial infarction, and 3 131 162 for pneumonia occurred. The overall post-discharge 30 day mortality was 8.7% for heart failure, 7.3% for acute myocardial infarction, and 8.4% for pneumonia. Risk adjusted mortality increased annually by 0.05% (95% confidence interval 0.02% to 0.08%) for heart failure, decreased by 0.06% (-0.09% to -0.04%) for acute myocardial infarction, and did not significantly change for pneumonia. Specifically, mortality increased for patients with heart failure who did not utilize any post-discharge acute care, increasing at a rate of 0.08% (0.05% to 0.12%) per year, exceeding the overall absolute annual increase in post-discharge mortality in heart failure, without an increase in mortality in observation units or the emergency department. Concurrent with a reduction in 30 day readmission rates, stays for observation and visits to the emergency department increased across all three conditions during and beyond the 30 day post-discharge period. Overall 30 day post-acute care utilization did not change significantly.

Conclusions: The only condition with increasing mortality through the study period was heart failure; the increase preceded the policy and was not present among patients who received emergency department or observation unit care without admission to hospital. During this period, the overall acute care utilization in the 30 days after discharge significantly decreased for heart failure and pneumonia, but not for acute myocardial infarction.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1136/bmj.l6831DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7190056PMC
January 2020

Association of Hospital Payment Profiles With Variation in 30-Day Medicare Cost for Inpatients With Heart Failure or Pneumonia.

JAMA Netw Open 2019 11 1;2(11):e1915604. Epub 2019 Nov 1.

Department of Health Care Policy, Harvard Medical School, Boston, Massachusetts.

Importance: Some uncertainty exists about whether hospital variations in cost are largely associated with differences in case mix.

Objective: To establish whether the same patients admitted with the same diagnosis (heart failure or pneumonia) at 2 different hospitals incur different costs associated with the hospital's Medicare payment profile.

Design, Setting, And Participants: This observational cohort study used Centers for Medicare & Medicaid Services (CMS) discharge data of patients with a principal diagnosis of heart failure (n = 1615) or pneumonia (n = 708) occurring between July 1, 2013, and June 30, 2016. Patients were individuals aged 65 years or older who were enrolled in Medicare fee-for-service Part A and Part B and were discharged from nonfederal, short-term, acute care or critical access hospitals in the United States. Data were analyzed from March 16, 2018, to September 25, 2019.

Main Outcomes And Measures: The CMS heart failure and pneumonia payment measure cohorts were divided into 2 random samples. In the first sample, hospitals were classified into payment quartiles for heart failure and pneumonia. In the second sample, patients with 2 admissions for heart failure or pneumonia, one in a lowest-quartile hospital and one in a highest-quartile hospital more than 1 month apart, were identified. Standardized Medicare payments for these patients were compared for the lowest- and the highest-quartile payment hospitals.

Results: The study sample included 1615 patients with heart failure (mean [SD] age, 78.7 [8.0] years; 819 [50.7%] male) and 708 with pneumonia (mean [SD] age, 78.3 [8.0] years; 401 [56.6%] male). The observed 30-day mortality rates for patients among lowest- compared with highest-payment hospitals were not significantly different. The median (interquartile range) hospital 30-day risk-standardized mortality rates were 8.1% (7.7%-8.5%) for heart failure and 11.3% (10.7%-12.1%) for pneumonia. The 30-day episode payment for hospitalization for the same patients at the lowest-payment hospitals was $2118 (95% CI, $1168-$3068; P < .001) lower for heart failure and $2907 (95% CI, $1760-$4054; P < .001) lower for pneumonia than at the highest-payment hospitals. More than half of the difference was associated with the payment during the index hospitalization ($1425 [95% CI, $695-$2154; P < .001] for heart failure and $1659 [95% CI, $731-$2588; P < .001] for pneumonia).

Conclusions And Relevance: This study found that the same Medicare beneficiaries who were admitted with the same diagnosis to hospitals with the highest payment profiles incurred higher costs than when they were admitted to hospitals with the lowest payment profiles. The findings suggest that variations in payments to hospitals are, at least in part, associated with the hospitals independently of non-time-varying patient characteristics.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1001/jamanetworkopen.2019.15604DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6902811PMC
November 2019

Substantial Differences Between Cohorts of Patients Hospitalized With Heart Failure in Canada and the United States.

JAMA Cardiol 2019 11;4(11):1178-1179

Center for Outcomes Research and Evaluation, Yale New Haven Hospital, New Haven, Connecticut.

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1001/jamacardio.2019.3314DOI Listing
November 2019

Development and Testing of Improved Models to Predict Payment Using Centers for Medicare & Medicaid Services Claims Data.

JAMA Netw Open 2019 08 2;2(8):e198406. Epub 2019 Aug 2.

Department of Biostatistics, Harvard T.H. Chan School of Public Health, Harvard University, Boston, Massachusetts.

Importance: Predicting payments for particular conditions or populations is essential for research, benchmarking, public reporting, and calculations for population-based programs. Centers for Medicare & Medicaid Services (CMS) models often group codes into disease categories, but using single, rather than grouped, diagnostic codes and leveraging present on admission (POA) codes may enhance these models.

Objective: To determine whether changes to the candidate variables in CMS models would improve risk models predicting patient total payment within 30 days of hospitalization for acute myocardial infarction (AMI), heart failure (HF), and pneumonia.

Design, Setting, And Participants: This comparative effectiveness research study used data from Medicare fee-for-service hospitalizations for AMI, HF, and pneumonia at acute care hospitals from July 1, 2013, through September 30, 2015. Payments across multiple care settings, services, and supplies were included and adjusted for geographic and policy variations, corrected for inflation, and winsorized. The same data source was used but varied for the candidate variables and their selection, and the method used by CMS for public reporting that used grouped codes was compared with variations that used POA codes and single diagnostic codes. Combinations of use of POA codes, separation of index admission diagnoses from those in the previous 12 months, and use of individual International Classification of Diseases, Ninth Revision, Clinical Modification codes instead of grouped diagnostic categories were tested. Data analysis was performed from December 4, 2017, to June 10, 2019.

Main Outcomes And Measures: The models' goodness of fit was compared using root mean square error (RMSE) and the McFadden pseudo R2.

Results: Among the 1 943 049 total hospitalizations of the study participants, 343 116 admissions were for AMI (52.5% male; 37.4% aged ≤74 years), 677 044 for HF (45.5% male; 25.9% aged ≤74 years), and 922 889 for pneumonia (46.4% male; 28.2% aged ≤74 years). The mean (SD) 30-day payment was $23 103 ($18 221) for AMI, $16 365 ($12 527) for HF, and $17 097 ($12 087) for pneumonia. Each incremental model change improved the pseudo R2 and RMSE. Incorporating all 3 changes improved the pseudo R2 of the patient-level models from 0.077 to 0.129 for AMI, from 0.042 to 0.129 for HF, and from 0.114 to 0.237 for pneumonia. Parallel improvements in RMSE were found for all 3 conditions.

Conclusions And Relevance: Leveraging POA codes, separating index from previous diagnoses, and using single diagnostic codes improved payment models. Better models can potentially improve research, benchmarking, public reporting, and calculations for population-based programs.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1001/jamanetworkopen.2019.8406DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6694388PMC
August 2019

Comparative Effectiveness of New Approaches to Improve Mortality Risk Models From Medicare Claims Data.

JAMA Netw Open 2019 07 3;2(7):e197314. Epub 2019 Jul 3.

Department of Biostatistics, Harvard T.H. Chan School of Public Health, Harvard University, Boston, Massachusetts.

Importance: Risk adjustment models using claims-based data are central in evaluating health care performance. Although US Centers for Medicare & Medicaid Services (CMS) models apply well-vetted statistical approaches, recent changes in the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) coding system and advances in computational capabilities may provide an opportunity for enhancement.

Objective: To examine whether changes using already available data would enhance risk models and yield greater discrimination in hospital-level performance measures.

Design, Setting, And Participants: This comparative effectiveness study used ICD-9-CM codes from all Medicare fee-for-service beneficiary claims for hospitalizations for acute myocardial infarction (AMI), heart failure (HF), or pneumonia among patients 65 years and older from July 1, 2013, through September 30, 2015. Changes to current CMS mortality risk models were applied incrementally to patient-level models, and the best model was tested on hospital performance measures to model 30-day mortality. Analyses were conducted from April 19, 2018, to September 19, 2018.

Main Outcomes And Measures: The main outcome was all-cause death within 30 days of hospitalization for AMI, HF, or pneumonia, examined using 3 changes to current CMS mortality risk models: (1) incorporating present on admission coding to better exclude potential complications of care, (2) separating index admission diagnoses from those of the 12-month history, and (3) using ungrouped ICD-9-CM codes.

Results: There were 361 175 hospital admissions (mean [SD] age, 78.6 [8.4] years; 189 225 [52.4%] men) for AMI, 716 790 hospital admissions (mean [SD] age, 81.1 [8.4] years; 326 825 [45.6%] men) for HF, and 988 225 hospital admissions (mean [SD] age, 80.7 [8.6] years; 460 761 [46.6%] men) for pneumonia during the study; mean 30-day mortality rates were 13.8% for AMI, 12.1% for HF, and 16.1% for pneumonia. Each change to the models was associated with incremental gains in C statistics. The best model, incorporating all changes, was associated with significantly improved patient-level C statistics, from 0.720 to 0.826 for AMI, 0.685 to 0.776 for HF, and 0.715 to 0.804 for pneumonia. Compared with current CMS models, the best model produced wider predicted probabilities with better calibration and Brier scores. Hospital risk-standardized mortality rates had wider distributions, with more hospitals identified as good or bad performance outliers.

Conclusions And Relevance: Incorporating present on admission coding and using ungrouped index and historical ICD-9-CM codes were associated with improved patient-level and hospital-level risk models for mortality compared with the current CMS models for all 3 conditions.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1001/jamanetworkopen.2019.7314DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6647547PMC
July 2019