Estimating state-specific population-based hospitalization rates from in-state hospital discharge data.

OBJECTIVE: To develop weights to estimate state population-based hospitalization rates for all residents of a state using only data from in-state hospitals which exclude residents treated in other states. DATA SOURCES AND STUDY SETTING: Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project (HCUP), State Inpatient Databases (SID), 2018-2019, 47 states+DC. STUDY DESIGN: We identified characteristics for patients hospitalized in each state differentiating movers (discharges for patients hospitalized outside state of residence) from stayers (discharges for patients hospitalized in state of residence) and created weights based on 2018 data informed by these characteristics. We calculated standard errors using a sampling framework and compared weight-based estimates against complete observed values for 2019. DATA COLLECTION/EXTRACTION METHODS: SID are based on administrative billing records collected by hospitals, shared with statewide data organizations, and provided to HCUP. PRINCIPAL FINDINGS: Of 34,186,766 discharged patients in 2018, 4.2% were movers. A higher share of movers (vs. stayers) lived in state border and rural counties; a lower share had discharges billed to Medicaid or were hospitalized for maternal/neonatal services. The difference between 2019 observed and estimated total discharges for all included states and DC was 9402 (mean absolute percentage error = 0.2%). We overestimated discharges with an expected payer of Medicaid, from the lowest income communities, and for maternal/neonatal care. We underestimated discharges with an expected payer of private insurance, from the highest income communities, and with injury diagnoses and surgical services. Estimates for most subsets were not within a 95% confidence interval, likely due to factors impossible to account for (e.g., hospital closures/openings, shifting consumer preferences). CONCLUSIONS: The weights offer a practical solution for researchers with access to only a single state's data to account for movers when calculating population-based hospitalization rates.

Effects of Hurricanes on Emergency Department Utilization: An Analysis Across 7 US Storms.

OBJECTIVE: Emergency departments (EDs) are critical sources of care after natural disasters such as hurricanes. Understanding the impact on ED utilization by subpopulation and proximity to the hurricane's path can inform emergency preparedness planning. This study examines changes in ED utilization for residents of 344 counties after the occurrence of 7 US hurricanes between 2005 and 2016. METHODS: This retrospective observational study used ED data from the Healthcare Cost and Utilization Project State Inpatient Databases and State Emergency Department Databases. ED utilization rates for weeks during and after hurricanes were compared with pre-hurricane rates, stratified by the proximity of the patient county to the hurricane path, age, and disease category. RESULTS: The overall population rate of weekly ED visits changed little post-hurricane, but rates by disease categories and age demonstrated varying results. Utilization rates for respiratory disorders exhibited the largest post-hurricane increase, particularly 2-3 weeks following the hurricane. The change in population rates by disease categories and age tended to be larger for people residing in counties closer to the hurricane path. CONCLUSIONS: Changes in ED utilization following hurricanes depend on disease categories, age, and proximity to the hurricane path. Emergency managers could incorporate these factors into their planning processes.

Readmissions Following Inpatient Treatment for Opioid-Related Conditions.

BACKGROUND: Previous research suggests that relatively few hospitalized patients with opioid-related conditions receive substance use treatment during their inpatient stay. Without treatment, these individuals may be more likely to have subsequent hospitalizations for continued opioid use disorder. OBJECTIVE: To evaluate the relationship between receipt of inpatient drug detoxification and/or rehabilitation services and subsequent opioid-related readmission. METHODS: This study used combined hospital inpatient discharge and emergency department visit data from the Agency for Healthcare Research and Quality's Healthcare Cost and Utilization Project. Our sample consisted of 329,037 patients from seven states with an opioid-related index hospitalization occurring between March 2010 and September 2013. Multivariate analysis was conducted to examine the relationship between opioid-related readmission and the receipt of inpatient drug detoxification and/or rehabilitation during the index visit. RESULTS: A relatively small percentage (19.4%) of patients with identified opioid-related conditions received treatment for drug use during their hospital inpatient stay. Patients who received drug rehabilitation, but not drug detoxification, during an opioid-related index hospitalization had lower odds of an opioid-related readmission within 90 days of discharge (odds ratio = 0.60, 95% confidence interval = 0.54-0.67) compared with patients with no inpatient drug detoxification or rehabilitation. Conclusions/Importance: A low percentage of patients receive inpatient services for drug use during an index stay involving an opioid-related diagnosis. Our findings indicate that receipt of drug rehabilitation services in acute care hospitals is associated with a lower 90-day readmission rate. Further research is needed to understand factors associated with the receipt of inpatient services and readmissions.

State variation in opioid treatment policies and opioid-related hospital readmissions.

BACKGROUND: State policy approaches designed to provide opioid treatment options have received significant attention in addressing the opioid epidemic in the United States. In particular, expanded availability of naloxone to reverse overdose, Good Samaritan laws intended to protect individuals who attempt to provide or obtain emergency services for someone experiencing an opioid overdose, and expanded coverage of medication-assisted treatment (MAT) for individuals with opioid abuse or dependence may help curtail hospital readmissions from opioids. The objective of this retrospective cohort study was to evaluate the association between the presence of state opioid treatment policies-naloxone standing orders, Good Samaritan laws, and Medicaid medication-assisted treatment (MAT) coverage-and opioid-related hospital readmissions. METHODS: We used 2013-2015 hospital inpatient discharge data from 13 states from the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project. We examined the relationship between state opioid treatment policies and 90-day opioid-related readmissions after a stay involving an opioid diagnosis. RESULTS: Our sample included 383,334 opioid-related index hospitalizations. Patients treated in states with naloxone standing-order policies at the time of the index stay had higher adjusted odds of an opioid-related readmission than did those treated in states without such policies; however, this relationship was not present in states with Good Samaritan laws. Medicaid methadone coverage was associated with higher odds of readmission among all insurance groups except Medicaid. Medicaid MAT coverage generosity was associated with higher odds of readmission among the Medicaid group but lower odds of readmission among the Medicare and privately insured groups. More comprehensive Medicaid coverage of substance use disorder treatment and a greater number of opioid treatment programs were associated with lower odds of readmission. CONCLUSIONS: Differences in index hospitalization rates suggest that states with opioid treatment policies had a higher level of need for opioid-related intervention, which also may account for higher rates of readmission. More research is needed to understand how these policies can be most effective in influencing acute care use.

Shifts in Alcohol-Related Diagnoses After the Introduction of International Classification of Diseases, Tenth Revision, Clinical Modification Coding in U.S. Hospitals: Implications for Epidemiologic Research.

BACKGROUND: In October 2015, the United States transitioned healthcare diagnosis codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), to the Tenth Revision (ICD-10-CM). Trend analyses of alcohol-related stays could show discontinuities solely from the change in classification systems. This study examined the impact of the ICD-10-CM coding system on estimates of hospital stays involving alcohol-related diagnoses. METHODS: This analysis used 2014 to 2017 administrative data from the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project State Inpatient Databases for 17 states. Quarterly ICD-9-CM data from second quarter 2014 through third quarter 2015 were concatenated with ICD-10-CM data from fourth quarter 2015 through first quarter 2017. Quarterly counts of alcohol-related stays were examined overall and then by 6 diagnostic subgroups: withdrawal, abuse, dependence, alcohol-induced mental disorders (AIMD), nonpsychiatric alcohol-induced disease, and intoxication or toxic effects. Within each group, we calculated the difference in the average number of stays between ICD-9-CM and ICD-10-CM coding periods. RESULTS: On average, the number of stays involving any alcohol-related diagnosis in the 6 quarters before and after the ICD-10-CM transition was stable. However, substantial shifts in stays occurred for alcohol abuse, AIMD, and intoxication or toxic effects. For example, the average quarterly number of stays involving AIMD was 170.7% higher in the ICD-10-CM period than in the ICD-9-CM period. This increase was driven in large part by 1 ICD-10-CM code, Alcohol use, unspecified with unspecified alcohol-induced disorder. CONCLUSIONS: Researchers conducting trend analyses of inpatient stays involving alcohol-related diagnoses should consider how ongoing modifications in the ICD-10-CM code system and coding guidelines might affect their work. An advisable approach for trend analyses across the ICD-10-CM transition is to aggregate diagnosis codes into broader, clinically meaningful groups-including a single global group that encompasses all alcohol-related stays-and then to select diagnostic groupings that minimize discontinuities between the 2 coding systems while providing useful information on this important indicator of population health.

Returns to Emergency Department, Observation, or Inpatient Care Within 30 Days After Hospitalization in 4 States, 2009 and 2010 Versus 2013 and 2014.

BACKGROUND: Nationally, readmissions have declined for acute myocardial infarction (AMI) and heart failure (HF) and risen slightly for pneumonia, but less is known about returns to the hospital for observation stays and emergency department (ED) visits. OBJECTIVE: To describe trends in rates of 30-day, all-cause, unplanned returns to the hospital, including returns for observation stays and ED visits. DESIGN: By using Healthcare Cost and Utilization Project data, we compared 210,007 index hospitalizations in 2009 and 2010 with 212,833 matched hospitalizations in 2013 and 2014. SETTING: Two hundred and one hospitals in Georgia, Nebraska, South Carolina, and Tennessee. PATIENTS: Adults with private insurance, Medicaid, or no insurance and seniors with Medicare who were hospitalized for AMI, HF, and pneumonia. MEASUREMENTS: Thirty-day hospital return rates for inpatient, observation, and ED visits. RESULTS: Return rates remained stable among adults with private insurance (15.1% vs 15.3%; P = 0.45) and declined modestly among seniors with Medicare (25.3% vs 25.0%; P = 0.04). Increases in observation and ED visits coincided with declines in readmissions (8.9% vs 8.2% for private insurance and 18.3% vs 16.9% for Medicare, both P ≤ 0.001). Return rates rose among patients with Medicaid (31.0% vs 32.1%; P = 0.04) and the uninsured (18.8% vs 20.1%; P = 0.004). Readmissions remained stable (18.7% for Medicaid and 9.5% for uninsured patients, both P > 0.75) while observation and ED visits increased. CONCLUSIONS: Total returns to the hospital are stable or rising, likely because of growth in observation and ED visits. Hospitalists' efforts to improve the quality and value of hospital care should consider observation and ED care.

Medicare Advantage and Traditional Medicare Hospitalization Intensity and Readmissions.

Medicare Advantage plans have incentives and tools to optimize patient care. Therefore, Medicare Advantage hospitalizations may have lower cost and higher quality than similar traditional Medicare hospitalizations. We applied a coarsened matching approach to 2013 Healthcare Cost and Utilization Project hospital discharge data from 22 states to compare hospital cost, length of stay, and readmissions for Traditional Medicare and Medicare Advantage. We found that Medicare Advantage hospitalizations were substantially less expensive and shorter for mental health stays but costlier and longer for injury and surgical stays. We found little difference in the cost and length of medical stays and in readmission rates. One explanation is that Medicare Advantage plans use outpatient settings for many patients with behavioral health conditions and for injury and surgical patients with less complex health needs. Alternatively, the observed differences in behavioral health cost and length of stay may represent skimping on appropriate care.

Trends in Opioid-related Inpatient Stays Shifted After the US Transitioned to ICD-10-CM Diagnosis Coding in 2015.

BACKGROUND: Trend analyses of opioid-related inpatient stays depend on the availability of comparable data over time. In October 2015, the US transitioned diagnosis coding from International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) to ICD-10-CM, increasing from ∼14,000 to 68,000 codes. This study examines how trend analyses of inpatient stays involving opioid diagnoses were affected by the transition to ICD-10-CM. SUBJECTS: Data are from Healthcare Cost and Utilization Project State Inpatient Databases for 14 states in 2015-2016, representing 26% of acute care inpatient discharges in the US. STUDY DESIGN: We examined changes in the number of opioid-related stays before, during, and after the transition to ICD-10-CM using quarterly ICD-9-CM data from 2015 and quarterly ICD-10-CM data from the fourth quarter of 2015 and the first 3 quarters of 2016. RESULTS: Overall, stays involving any opioid-related diagnosis increased by 14.1% during the ICD transition-which was preceded by a much lower 5.0% average quarterly increase before the transition and followed by a 3.5% average increase after the transition. In stratified analysis, stays involving adverse effects of opioids in therapeutic use showed the largest increase (63.2%) during the transition, whereas stays involving abuse and poisoning diagnoses decreased by 21.1% and 12.4%, respectively. CONCLUSIONS: The sharp increase in opioid-related stays overall during the transition to ICD-10-CM may indicate that the new classification system is capturing stays that were missed by ICD-9-CM data. Estimates of stays involving other diagnoses may also be affected, and analysts should assess potential discontinuities in trends across the ICD transition.

Racial Disparities in Sepsis-Related In-Hospital Mortality: Using a Broad Case Capture Method and Multivariate Controls for Clinical and Hospital Variables, 2004-2013.

OBJECTIVES: As sepsis hospitalizations have increased, in-hospital sepsis deaths have declined. However, reported rates may remain higher among racial/ethnic minorities. Most previous studies have adjusted primarily for age and sex. The effect of other patient and hospital characteristics on disparities in sepsis mortality is not yet well-known. Furthermore, coding practices in claims data may influence findings. The objective of this study was to use a broad method of capturing sepsis cases to estimate 2004-2013 trends in risk-adjusted in-hospital sepsis mortality rates by race/ethnicity to inform efforts to reduce disparities in sepsis deaths. DESIGN: Retrospective, repeated cross-sectional study. SETTING: Acute care hospitals in the Healthcare Cost and Utilization Project State Inpatient Databases for 18 states with consistent race/ethnicity reporting. PATIENTS: Patients diagnosed with septicemia, sepsis, organ dysfunction plus infection, severe sepsis, or septic shock. MEASUREMENTS AND MAIN RESULTS: In-hospital sepsis mortality rates adjusted for patient and hospital factors by race/ethnicity were calculated. From 2004 to 2013, sepsis hospitalizations for all racial/ethnic groups increased, and mortality rates decreased by 5-7% annually. Mortality rates adjusted for patient characteristics were higher for all minority groups than for white patients. After adjusting for hospital characteristics, sepsis mortality rates in 2013 were similar for white (92.0 per 1,000 sepsis hospitalizations), black (94.0), and Hispanic (93.5) patients but remained elevated for Asian/Pacific Islander (106.4) and "other" (104.7; p < 0.001) racial/ethnic patients. CONCLUSIONS: Our results indicate that hospital characteristics contribute to higher rates of sepsis mortality for blacks and Hispanics. These findings underscore the importance of ensuring that improved sepsis identification and management is implemented across all hospitals, especially those serving diverse populations.

The Shifting Landscape in Utilization of Inpatient, Observation, and Emergency Department Services Across Payers.

Recent policies by public and private payers have increased incentives to reduce hospital admissions. Using data from four states from the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project, this study compared the payer-specific population-based rates of adults using inpatient, observation, and emergency department (ED) services for 10 common medical conditions in 2009 and in 2013. Patients had an expected primary payer of private insurance, Medicare, Medicaid, or no insurance. Across all four payer populations, inpatient admissions declined, and care shifted toward treat-and-release observation stays and ED visits. The percentage of hospitalizations that began with an observation stay increased. Implications for quality of care and costs to patients warrant further examination. Journal of Hospital Medicine 2017;12:443-446.

Managed care and inpatient mortality in adults: effect of primary payer.

BACKGROUND: Because managed care is increasingly prevalent in health care finance and delivery, it is important to ascertain its effects on health care quality relative to that of fee-for-service plans. Some stakeholders are concerned that basing gatekeeping, provider selection, and utilization management on cost may lower quality of care. To date, research on this topic has been inconclusive, largely because of variation in research methods and covariates. Patient age has been the only consistently evaluated outcome predictor. This study provides a comprehensive assessment of the association between managed care and inpatient mortality for Medicare and privately insured patients. METHODS: A cross-sectional design was used to examine the association between managed care and inpatient mortality for four common inpatient conditions. Data from the 2009 Healthcare Cost and Utilization Project State Inpatient Databases for 11 states were linked to data from the American Hospital Association Annual Survey Database. Hospital discharges were categorized as managed care or fee for service. A phased approach to multivariate logistic modeling examined the likelihood of inpatient mortality when adjusting for individual patient and hospital characteristics and for county fixed effects. RESULTS: Results showed different effects of managed care for Medicare and privately insured patients. Privately insured patients in managed care had an advantage over their fee-for-service counterparts in inpatient mortality for acute myocardial infarction, stroke, pneumonia, and congestive heart failure; no such advantage was found for the Medicare managed care population. To the extent that the study showed a protective effect of privately insured managed care, it was driven by individuals aged 65 years and older, who had consistently better outcomes than their non-managed care counterparts. CONCLUSIONS: Privately insured patients in managed care plans, especially older adults, had better outcomes than those in fee-for-service plans. Patients in Medicare managed care had outcomes similar to those in Medicare FFS. Additional research is needed to understand the role of patient selection, hospital quality, and differences among county populations in the decreased odds of inpatient mortality among patients in private managed care and to determine why this result does not hold for Medicare.

Differences in Use of High-quality and Low-quality Hospitals Among Working-age Individuals by Insurance Type.

BACKGROUND: Research suggests that individuals with Medicaid or no insurance receive fewer evidence-based treatments and have worse outcomes than those with private insurance for a broad range of conditions. These differences may be due to patients' receiving care in hospitals of different quality. RESEARCH DESIGN: We used the Healthcare Cost and Utilization Project State Inpatient Databases 2009-2010 data to identify patients aged 18-64 years with private insurance, Medicaid, or no insurance who were hospitalized with acute myocardial infarction, heart failure, pneumonia, stroke, or gastrointestinal hemorrhage. Multinomial logit regressions estimated the probability of admissions to hospitals classified as high, medium, or low quality on the basis of risk-adjusted, in-hospital mortality. RESULTS: Compared with patients who have private insurance, those with Medicaid or no insurance were more likely to be minorities and to reside in areas with low-socioeconomic status. The probability of admission to high-quality hospitals was similar for patients with Medicaid (23.3%) and private insurance (23.0%) but was significantly lower for patients without insurance (19.8%, P<0.01) compared with the other 2 insurance groups. Accounting for demographic, socioeconomic, and clinical characteristics did not influence the results. CONCLUSIONS: Previously noted disparities in hospital quality of care for Medicaid recipients are not explained by differences in the quality of hospitals they use. Patients without insurance have lower use of high-quality hospitals, a finding that needs exploration with data after 2013 in light of the Affordable Care Act, which is designed to improve access to medical care for patients without insurance.

Understanding Racial and Ethnic Disparities in Postsurgical Complications Occurring in U.S. Hospitals.

OBJECTIVE: To examine the role of patient, hospital, and community characteristics on racial and ethnic disparities in in-hospital postsurgical complications. DATA SOURCES: Healthcare Cost and Utilization Project, 2011 State Inpatient Databases; American Hospital Association Annual Survey of Hospitals; Area Health Resources Files; Centers for Medicare & Medicaid Services Hospital Compare database. METHODS: Nonlinear hierarchical modeling was conducted to examine the odds of patients experiencing any in-hospital postsurgical complication, as defined by Agency for Healthcare Research and Quality Patient Safety Indicators. PRINCIPAL FINDINGS: A total of 5,474,067 inpatient surgical discharges were assessed using multivariable logistic regression. Clinical risk, payer coverage, and community-level characteristics (especially income) completely attenuated the effect of race on the odds of postsurgical complications. Patients without private insurance were 30 to 50 percent more likely to have a complication; patients from low-income communities were nearly 12 percent more likely to experience a complication. Private, not-for-profit hospitals in small metropolitan or micropolitan areas and higher nurse-to-patient ratios led to fewer postsurgical complications. CONCLUSIONS: Race does not appear to be an important determinant of in-hospital postsurgical complications, but insurance and community characteristics have an effect. A population-based approach that includes improving the socioeconomic context may help reduce disparities in these outcomes.

Age-related Disparities in Trauma Center Access for Severe Head Injuries Following the Release of the Updated Field Triage Guidelines.

OBJECTIVE: In 2006, the American College of Surgeons' Committee on Trauma and the Centers for Disease Control and Prevention released field triage guidelines with special consideration for older adults. Additional considerations for direct transport to a Level I or II trauma center (TC) were added in 2011, reflecting perceived undertriage to TCs for older adults. We examined whether age-based disparities in TC care for severe head injury decreased following introduction of the 2011 revisions. METHODS: A pre-post design analyzing the 2009 and 2012 Healthcare Cost and Utilization Project State Emergency Department Databases and State Inpatient Databases with multivariable logistic regressions considered changes in 1) the trauma designation of the emergency department where treatment was initiated and 2) transfer to a TC following initial treatment at a non-TC. RESULTS: Compared with adults aged 18 to 44 years, after multivariable adjustment, in both years TC care was less likely for adults aged 45 to 64 years (odds ratio [OR] = 0.76 in 2009 and 0.74 in 2012), aged 65 to 84 years (OR = 0.61 and 0.59), and aged 85+ years (OR = 0.53 and 0.56). Between 2009 and 2012, the likelihood of TC care increased for all age groups, with the largest increase among those aged 85+ years (OR = 1.18), which was statistically different (p = 0.02) from the increase among adults aged 18 to 44 years (OR = 1.12). The analysis of transfers yielded similar results. CONCLUSIONS: Although patterns of increased TC treatment for all groups with severe head trauma indicate improvements, age-based disparities persisted.

Impact of Race/Ethnicity and Socioeconomic Status on Risk-Adjusted Hospital Readmission Rates Following Hip and Knee Arthroplasty.

BACKGROUND: Readmission rates following total hip arthroplasty (THA) and total knee arthroplasty (TKA) are increasingly used to measure hospital performance. Readmission rates that are not adjusted for race/ethnicity and socioeconomic status, patient risk factors beyond a hospital's control, may not accurately reflect a hospital's performance. In this study, we examined the extent to which risk-adjusting for race/ethnicity and socioeconomic status affected hospital performance in terms of readmission rates following THA and TKA. METHODS: We calculated 2 sets of risk-adjusted readmission rates by (1) using the Centers for Medicare & Medicaid Services standard risk-adjustment algorithm that incorporates patient age, sex, comorbidities, and hospital effects and (2) adding race/ethnicity and socioeconomic status to the model. Using data from the Healthcare Cost and Utilization Project, 2011 State Inpatient Databases, we compared the relative performances of 1,194 hospitals across the 2 methods. RESULTS: Addition of race/ethnicity and socioeconomic status to the risk-adjustment algorithm resulted in (1) little or no change in the risk-adjusted readmission rates at nearly all hospitals; (2) no change in the designation of the readmission rate as better, worse, or not different from the population mean at >99% of the hospitals; and (3) no change in the excess readmission ratio at >97% of the hospitals. CONCLUSIONS: Inclusion of race/ethnicity and socioeconomic status in the risk-adjustment algorithm led to a relative-performance change in readmission rates following THA and TKA at <3% of the hospitals. We believe that policymakers and payers should consider this result when deciding whether to include race/ethnicity and socioeconomic status in risk-adjusted THA and TKA readmission rates used for hospital accountability, payment, and public reporting. LEVEL OF EVIDENCE: Prognostic Level III. See instructions for Authors for a complete description of levels of evidence.

Length of stay in EDs: variation across classifications of clinical condition and patient discharge disposition.

STUDY OBJECTIVE: Duration of a stay in an emergency department (ED) is considered a measure of quality, but current measures average lengths of stay across all conditions. Previous research on ED length of stay has been limited to a single condition or a few hospitals. We use a census of one state's data to measure length of ED stays by patients' conditions and dispositions and explore differences between means and medians as quality metrics. METHODS: The data source was the Healthcare Cost and Utilization Project 2011 State Emergency Department Databases and State Inpatient Databases for Florida. Florida is unique in collecting ED length of stay for both released and admitted patients. Clinical Classifications Software was used to group visits based on first-listed International Classification of Disease, Ninth Edition, Clinical Modification, diagnoses. RESULTS: For the 10 most common diagnoses, patients with relatively minor injuries typically required the shortest mean stay (3 hours or less); conditions resulting in admission or transfer tended to be more serious, resulting in longer stays. Patients requiring the longest stays, by disposition, had discharge diagnoses of nonspecific chest pain (mean 7.4 hours among discharged patients), urinary tract infections (4.8 hours among admissions), and schizophrenia (9.6 hours among transfers) among the top 10 diagnoses. CONCLUSION: Emergency department length of stay as a measure of ED quality should take into account the considerable variation by condition and disposition of the patient. Emergency department length of stay measurement could be improved in the United States by standardizing its definition; distinguishing visits involving treatment, observation, and boarding; and incorporating more distributional information.

Emergency department transfers and transfer relationships in United States hospitals.

OBJECTIVES: The objective was to describe transfers out of hospital-based emergency departments (EDs) in the United States and to identify different characteristics of sending and receiving hospitals, travel distance during transfer, disposition on arrival to the second hospital, and median number of transfer partners among sending hospitals. METHODS: Emergency department records were linked at transferring hospitals to ED and inpatient records at receiving hospitals in nine U.S. states using the 2010 Healthcare Cost and Utilization Project (HCUP) State Emergency Department Databases and State Inpatient Databases, the American Hospital Association Annual Survey, and the Trauma Information Exchange Program. Using the Clinical Classification Software (CCS) to categorize conditions, the 50 disease categories with the highest transfer rates were studied, and these were then placed into nine clinical groups. Records were included where both sending and receiving records were available; these data were tabulated to describe ED transfer patterns, hospital-to-hospital distances, final patient disposition, and number of transfer partners. RESULTS: A total of 97,021 ED transfer encounters were included in the analysis from the 50 highest transfer rate disease categories. Among these, transfer rates ranged from 1% to 13%. Circulatory conditions made up about half of all transfers. Receiving hospitals were more likely to be nonprofit, teaching, trauma, and urban and have more beds with greater specialty coverage and more advanced diagnostic and therapeutic resources. The median transfer distance was 23 miles, with 25% traveling more than 40 to 50 miles. About 8% of transferred encounters were discharged from the second ED, but that varied from 0.6% to 53% across the 50 conditions. Sending hospitals had a median of seven transfer partners across all conditions and between one and four per clinical group. CONCLUSIONS: Among high-transfer conditions in U.S. EDs, patients are often transferred great distances, more commonly to large teaching hospitals with greater resources. The large number of transfer partners indicates a possible lack of stable transfer relationships between U.S. hospitals.

The transfer instability index: a novel metric of emergency department transfer relationships.

OBJECTIVES: In this study, the objective was to characterize emergency department (ED) transfer relationships and study the factors that predict the stability of those relationships. A metric is derived for ED transfer relationships that may be useful in assessing emergency care regionalization and as a resource for future emergency medicine research. METHODS: Emergency department records at transferring hospitals were linked to ED and inpatient records at receiving hospitals in nine U.S. states using the 2010 Healthcare Cost and Utilization Project State Emergency Department Databases and State Inpatient Databases, the American Hospital Association Annual Survey, and the Trauma Information Exchange Program. Using the Clinical Classification Software to categorize conditions, high transfer rate conditions were placed into nine clinical groups. The authors created a new measure, the "transfer instability index," which estimates the effective number of "transfer partners" for each sending ED: this is designed to measure the stability of outgoing transfer relationships, where higher values of the index indicate less stable relationships. The index provides a measure of how many hospitals a transferring hospital sends its patients to (weighted by how often each transfer partner is used). Regression was used to analyze factors associated with higher values of the index. RESULTS: Sending hospitals had a median of 3.5 effective transfer partners across all conditions. The calculated transfer instability indices varied from 1 to 2.4 across disease categories. In general, higher index values were associated with treating a higher proportion of publicly insured patients: 10 and 12% increases in the Medicare and Medicaid share of ED encounters, respectively, were associated with 10 and 14% increases in the effective number of transfer partners. This public insurance effect held while studying all conditions together as well as within individual disease categories, such as cardiac, neurologic, and traumatic conditions. CONCLUSIONS: United States EDs that transfer patients to other hospitals often have multiple transfer partners. The stability of the transfer relationship, assessed by the transfer instability index, differs by condition. Less stable transfer relationships (i.e., hospitals with greater numbers of transfer partners) were more common in EDs with higher proportions of publicly insured patients.