Differences in personal characteristics and health outcomes between ambulatory and non-ambulatory adults with traumatic spinal cord injury.

OBJECTIVE: To identify differences in personal characteristics, health outcomes, and hospital utilization as a function of ambulatory status among adults with chronic SCI. DESIGN: Prospective cohort study linked to state administrative billing data. SETTING: Population-based SCI Registry from the Southeastern United States. PARTICIPANTS: 1,051 adults (>18 years old) with chronic (>1-year), traumatic SCI. OUTCOME MEASURES: The self-report assessment (SRA) included demographic, injury and disability characteristics, health status, psychological and behavioral factors, and participation and quality of life (QOL) variables. We linked cases to administrative billing data to assess hospital utilization, including Emergency Department (ED) visits and inpatient (IP) admissions (through the ED and direct IP) in non-federal state hospitals within the year following the SRA. RESULTS: There were 706 ambulatory and 345 non-ambulatory participants. We found significant differences across all sets of factors and significant differences in hospital utilization metrics. Ambulatory adults had fewer ED visits (36% vs 44%), IP admissions through the ED (11% vs 25%) and IP only admissions (9% vs 19%) and spent fewer days in the hospital for both admissions through the ED (0.9 vs 4.6 days) and IP only admissions (0.7 vs 3.1 days). They also reported having fewer past year ED visits (44% vs 62%) and IP admissions (34% vs 52%). CONCLUSIONS: We identified differences in personal characteristics, ED visits and IP admissions between ambulatory and non-ambulatory adults with SCI, providing a better understanding of the characteristics of those with SCI. The findings suggest the need for separate analyses based on ambulatory status when assessing long-term health outcomes including hospital utilization.

Assessment of high-risk opioid use metrics among individuals with spinal cord injury: A brief report.

OBJECTIVE: To examine high-risk opioid prescription metrics among individuals with chronic spinal cord injury (SCI) living in South Carolina. DESIGN: Cohort Study. SETTING: Two statewide population-based databases, an SCI Surveillance Registry and state prescription drug monitoring program (PDMP). PARTICIPANTS: Linked data was obtained for 503 individuals with chronic (>1year-post injury) SCI who were injured in 2013 or 2014 and who survived at least 3 years post-injury. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE(S): Opioid prescription metrics were obtained from the PDMP. Fill data during the period between January 1, 2014 and December 31, 2017 were analyzed to assess high-risk opioid use. Outcomes included: percentage of individuals prescribed chronic opioids, high-dose chronic opioid therapy (daily morphine milligram equivalents (MME) ≥50 and ≥90), and chronic concurrent opioids and benzodiazepines, sedatives, or hypnotics (BSH). RESULTS: Over half (53%) of the individuals filled an opioid in years 2-3 after injury. Of those, 38% had a concurrent BSH fill during the study period, 76% of which were for benzodiazepines. In any given quarter over the two-year timeframe, over half of the opioid prescriptions were for 60 days or more (chronic opioid prescriptions). Of those, roughly 40% of the individuals had high-dose chronic opioid prescriptions ≥50 MME/d and 25% were ≥90 MME/d. Over 33% had a concurrent BSH prescription for ≥60 days. CONCLUSIONS: While the number of individuals receiving high-risk opioid prescriptions may not be large, it is a concerning number of prescriptions. The findings suggest a need for more cautious opioid prescribing and monitoring of high-risk use in adults with chronic SCI.

Rehospitalization During the First 5 Years After the Onset of Traumatic Spinal Cord Injury: A Population-Based Study Using Administrative Billing Records.

OBJECTIVE: Our objective was to identify the number, length of stay, and charges of rehospitalizations during the subsequent 5 years after discharge from the initial hospitalization by using administrative billing records from a population-based cohort with spinal cord injury (SCI) in the southeastern United States. DESIGN: Analysis of administrative billing data. SETTING: State-based surveillance data analyzed by an academic medical center in the southeastern United States. PARTICIPANTS: A total of 1872 individuals (N=1872) from a state-based surveillance system in the southeastern United States whose onset was between January 1, 1998, and January 1, 2010. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: The outcome measures were the number of rehospitalization episodes, length of stay, and total hospital charges for each episode of rehospitalization. RESULTS: Seventy percent of participants were rehospitalized during the first 5 years after initial discharge, and the highest rehospitalization rates were in the first year (54%), being relatively stable in years 2-5 (21%-22%). Adjusted to 2019 US dollars, the average total rehospitalization charges were $214,716 per person during the 5 years. Participants who could walk independently had fewer rehospitalizations, fewer rehospitalization days, and less rehospitalization charges than the nonambulatory participants. College education was also associated with less rehospitalization charges. CONCLUSIONS: Rehospitalization is a significant cost after SCI. Further longitudinal study on the population cohorts and billing data are needed to quantify these changes over time.

The primary and secondary causes of hospitalizations during the first five years after spinal cord injury.

STUDY DESIGN: Cohort study. Retrospective analysis of uniform billing discharge data (UB-04). OBJECTIVES: To compare and contrast the primary and secondary causes of hospitalization by type of admission, emergency department (ED) versus inpatient only (IP), during the first five years after the traumatic spinal cord injury (SCI). SETTING: Academic Medical University in the Southeastern USA. METHODS: At total of 2569 adults with traumatic SCI were identified from a population-based registry and matched to billing data. The main outcome measures were primary and secondary diagnoses associated with hospital admissions in non-federal, state hospitals. RESULTS: Overall, there were 9733 hospital admissions in the five years after SCI onset, not including the initial hospitalization; 53% were admissions through the ED. The primary causes of hospitalizations after SCI varied by year post injury and admission type (ED versus IP). The top 15 secondary diagnoses included several secondary health conditions associated with SCI, as well as chronic health conditions. CONCLUSIONS: Rehabilitation diagnoses were much more prominent during the first year, compared with subsequent years. Septicemia was the leading cause of admissions through the ED, whereas chronic ulcers of the skin were prominent for IP only admissions. This is consistent with the acute nature of septicemia compared with more planned hospitalization for rehabilitation and skin ulcers. These conditions should be targeted for prevention strategies that include patient/family education and early and appropriate access to primary care.

Opioid Use Among Individuals With Spinal Cord Injury: Prevalence Estimates Based on State Prescription Drug Monitoring Program Data.

OBJECTIVE: To identify the prevalence of opioid use in individuals with chronic spinal cord injury (SCI) living in South Carolina. DESIGN: Cohort study. SETTING: Data from 2 statewide population-based databases, an SCI Registry and the state prescription drug monitoring program, were linked and analyzed. PARTICIPANTS: The study included individuals (N=503) with chronic (>1y) SCI who were injured between 2013 and 2014 in South Carolina and who survived at least 3 years postinjury. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Filled opioid prescriptions over a 2-year period (months 13-36 after injury). The main outcomes were total number of days with an opioid prescription over the 2-year period, length of coverage period [(final day of prescription coverage+the days supplied)-first day of prescription coverage], average daily morphine milligram equivalents (MME) over the coverage period, and concurrent days covered by an opioid and a prescription for benzodiazepines, sedatives, or hypnotics. RESULTS: A total of 53.5% of the cohort (269 individuals) filled at least 1 opioid prescription during their second or third year after SCI. In total, there were 3386 opioid fills during the 2-year study. On average, the total number of opioid prescription days was 293±367. The average coverage period was 389±290 days, and the average daily MME during the coverage period was 41±70 MME. Of those who filled an opioid prescription, 23% had high-risk fills (>50 MME), and 38% had concurrent prescriptions for benzodiazepines, sedatives, or hypnotics. CONCLUSIONS: The prevalence of opioid use was high among individuals with chronic SCI, exceeding rates observed in the general population. Also concerning were the rates of high-risk fills, based on average daily MME and concurrent benzodiazepine, sedative, or hypnotic prescriptions. These findings, taken together with those of earlier studies, should be used by providers to assess and monitor opioid use, decrease concurrent high-risk medication use, and attenuate the risk of adverse outcomes.

The association between participation and quality of life indicators with hospitalizations in ambulatory adults with spinal cord injury.

STUDY DESIGN: Statistical modeling of self-report assessments (SRA) as predictors of future hospitalizations, measured by administrative billing data. OBJECTIVES: To examine the relationships between self-reported participation and quality of life (QOL) indicators and future hospital admissions among ambulatory adults with chronic spinal cord injury (SCI). SETTING: Data were collected from participants living in and utilizing hospitals in the state of South Carolina. METHODS: Participants were identified through the South Carolina SCI Surveillance System Registry. Between 2011 and 2015, 615 ambulatory adults (>18 years old) with chronic (>1-year), traumatic SCI completed mailed SRA. Participant socio-demographic, injury, health, participation, and QOL indicators were assessed using self-report data. Administrative billing data were used to measure hospital utilization in nonfederal, South Carolina hospitals in the year following the SRA. RESULTS: Prior year discharges, current pressure ulcers, number of chronic conditions, walking 150 feet more often (never, less than once per week, at least once per week, once or twice per day, or several times per day), and greater home life satisfaction were associated with an increased risk of subsequent hospitalization. Walking 10 feet more frequently and greater global satisfaction were associated with a decreased risk of hospital admission. CONCLUSIONS: Specific participation and QOL items may increase the risk of hospitalization in ambulatory adults with SCI. Further study is necessary to understand better the relationships between walking distance and frequency, home life and global satisfaction, and inpatient admissions.

Correction: Prevalence of chronic health conditions and hospital utilization in adults with spinal cord injury: an analysis of self-report and South Carolina administrative billing data.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Concentration of Costs Among High Utilizers of Health Care Services Over the First 10 Years After Spinal Cord Injury Rehabilitation: A Population-based Study.

OBJECTIVE: The purpose of this study was to (1) categorize individuals into high, medium, and low utilizers of health care services over a 10-year period after the onset of spinal cord injury (SCI) and (2) identify the pattern of causes of hospitalizations and the characteristics associated with high utilization. DESIGN: Retrospective analysis of self-report assessment linked to administrative data. SETTING: Data were collected from participants living in and utilizing hospitals in the state of South Carolina. PARTICIPANTS: Adult participants with traumatic SCI were identified through a state SCI Surveillance System Registry, a population-based system capturing all incident cases treated in nonfederal facilities. Among 963 participants who completed self-report assessments, we matched those with a minimum of 10 years of administrative records for a final sample of 303 participants (N=303). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Costs related to health care utilization for emergency department visits and hospitalizations, as measured operationally by hospital charges at full and established rates; causes of hospitalizations RESULTS: Over two-thirds of the total $49.4 million in charges for hospitalization over the 10-year timeframe (69%) occurred among 16.5% of the cohort (high utilizers), whereas those in the low utilizer group comprised 53% of the cohort with only 3.5% of the charges. The primary diagnoses were septicemia (50%), other urinary tract disorder (48%), mechanical complication of device, implant, or graft (48%), and chronic ulcer of skin (40%). Primary diagnoses were frequently accompanied by secondary diagnoses, indicating the co-occurrence of multiple secondary health conditions. High utilizers were more likely to be male, minority, have a severe SCI, have reported frequent pressure ulcers and have income of less than $35,000 per year. CONCLUSIONS: The high cost of chronic health care utilization over a 10-year timeframe was concentrated in a relatively small portion of the SCI population who have survived more than a decade after SCI onset.

Prevalence of chronic health conditions and hospital utilization in adults with spinal cord injury: an analysis of self-report and South Carolina administrative billing data.

STUDY DESIGN: Retrospective analysis of self-report and administrative billing data. OBJECTIVES: (1) Identify the self-reported prevalence of seven chronic health conditions (CHCs) in adults with chronic, traumatic spinal cord injury (SCI), (2) Examine the relationships between the presence of CHC with future hospital admissions and total number of inpatient days and (3) identify predictors of utilization. SETTING: Data were collected from participants living in and utilizing hospitals in South Carolina, USA. METHODS: Participants were identified through the South Carolina SCI Surveillance System Registry. Between 2010 and 2013, 963 adults ( > 18 years old) with chronic ( > 1-year), traumatic SCI completed self-report assessments (SRAs); this analysis includes data from 787 individuals. The presence/absence of the seven CHC was assessed using self-report data. Administrative billing data were used to assess hospital utilization in non-federal, South Carolina hospitals in the year following the SRA. RESULTS: In all, 40.5% reported no CHC; 23.4% reported one CHC and 36.1% reported having two or more CHC. The most commonly reported CHCs were hypertension (43.1%), high cholesterol (32.2%) and diabetes (15.8%). In total, 59% had at least one hospital admission in the year following the SRA (mean 3 ± 5; range 0-45; median = 1). The mean total inpatient days was 15.7 ± 43 days (range 0-365; median = 1). Predictors of hospital admission included CHC, pressure sores, education, prior hospitalization and injury severity. With the exception of CHC, each was also associated with total number of inpatient days. CONCLUSIONS: CHC are prevalent and associated with hospital admissions in adults with chronic, traumatic SCI.

Comparison of Rates of Hospitalization and Emergency Department Visits Using Self-Report and South Carolina Administrative Billing Data Among a Population-Based Cohort With Spinal Cord Injury.

OBJECTIVE: To compare self-report and South Carolina administrative billing data documentation of emergency department (ED) visits and hospitalizations in the past 12 months among a population-based cohort of persons with spinal cord injury (SCI). DESIGN: Cross-sectional study. SETTING: SCI surveillance system in South Carolina. PARTICIPANTS: Persons (N=605) sustaining a traumatic SCI between January 1, 1998 and December 31, 2011 in South Carolina who, at the time of study assessment, were adults, were >1 years postinjury, and had not made a complete recovery. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Hospitalizations and ED visits in the past 12 months. RESULTS: There was a significantly higher rate of reporting >1 hospitalizations in the past year for self-report (36%) as compared with South Carolina administrative billing data (26%) (P<.001), but not for >1 ED visits (48% vs 45%; P=.11). Decreased physical health and increased injury severity were associated with higher reporting rates of hospitalization. Physical health and injury severity were predictive of both self-report and South Carolina administrative billing data of hospitalizations, whereas years postinjury and race were also predictors of South Carolina administrative billing data hospitalizations. CONCLUSIONS: Our comparison of self-report and South Carolina administrative billing data hospitalizations and ED visits showed a significantly higher rate of reporting of hospitalizations using self-report, specifically among those with poor physical health and higher injury severity. Future work should look at different ways of asking about health care utilization and compare with South Carolina administrative billing data documentation to identify the best ways to assess through self-report.

The relationship between left ventricular scar and ventricular repolarization in patients with coronary artery disease: insights from late gadolinium enhancement magnetic resonance imaging.

AIMS: The markers of ventricular repolarization corrected QT interval (QTc), QT dispersion (QTD) and Tpeak-to-Tend interval (Tpeak-end) have shown an association with sudden cardiac death (SCD) in the general population. However, their mechanistic relationship with SCD is unclear. The study aim was to evaluate the relationship between QTc, QTD, and Tpeak-end, and the extent and distribution of left ventricular (LV) scar in patients with coronary artery disease at high SCD risk. METHODS AND RESULTS: We included 64 consecutive implantable cardioverter defibrillator (ICD) recipients (66 ± 11 years, 80% male, median left ventricular ejection fraction 30%) who had undergone late gadolinium enhancement cardiac magnetic resonance (CMR) imaging prior to device implantation over 4 years. Scar was quantified using the CMR images and characterized in terms of percent LV scar and number of LV segments with subendocardial/transmural scar. Repolarization parameters were measured on an electrocardiogram performed prior to ICD implantation. After adjustment for potential confounders there was a strong association between the number of limited subendocardial (1-25% transmurality) scar segments and QTc (P = 0.003), QTD (P = 0.002), and Tpeak-end (P = 0.008). However, there was no association between the repolarization parameters and percent LV scar or the amount of transmural scar. During a mean follow-up of 19 ± 10 months 19 (30%) patients received appropriate ICD therapy, but none of the repolarization parameters were associated with its occurrence. CONCLUSION: In this pilot study there was a strong association between limited subendocardial LV scar and prolonged QTc, QTD, and Tpeak-end. However, there was no association between any of these repolarization markers and the delivery of appropriate ICD therapy.

Left ventricular scar burden specifies the potential for ventricular arrhythmogenesis: an LGE-CMR study.

INTRODUCTION: The extent of left ventricular (LV) scar, characterized by late gadolinium enhancement cardiac MRI (LGE-CMR), has been shown to predict the occurrence of ventricular arrhythmias in implantable cardioverter defibrillator (ICD) recipients. However, the specificity of LGE-CMR for sudden cardiac death (SCD) versus non-SCD is unclear. The aim of this retrospective, observational study was to evaluate this relationship in a cohort of ICD recipients. METHODS AND RESULTS: We included consecutive patients who had undergone LGE-CMR before ICD implantation over a 4-year period (2006-2009). Scar (defined as myocardium with a signal intensity ≥50% of the maximum in scar tissue) was characterized in terms of percent scar and number of transmural LV scar segments in a 17-segment model. The endpoints were appropriate ICD therapy and all-cause mortality. Sixty-four patients (average age 66 ± 11 years, 51 male, median LVEF 30%) were included. During 42 ± 13 months follow-up, appropriate ICD therapy occurred in 28 patients (44%), and 14 patients (22%) died. Number of transmural scar segments (P = 0.005) and percentage LV scar (P = 0.03) were both significantly associated with appropriate ICD therapy. However, neither number of transmural scar segments (P = 0.32) or percent LV scar (P = 0.59) was significantly associated with all-cause mortality. CONCLUSION: In this observational study, in medium-term follow-up, the extent of LV scar characterized by LGE-CMR was strongly associated with the occurrence of spontaneous ventricular arrhythmias but not all-cause mortality. We hypothesize that scar quantification by LGE-CMR may be more specific for SCD than non-SCD, and may prove a valuable tool for the selection of patients for ICD therapy.

The extent of left ventricular scar quantified by late gadolinium enhancement MRI is associated with spontaneous ventricular arrhythmias in patients with coronary artery disease and implantable cardioverter-defibrillators.

BACKGROUND: Characterization of sudden cardiac death (SCD) risk remains a challenge in the application of implantable cardioverter-defibrillator (ICD) therapy. Late gadolinium enhancement cardiac MRI (LGE-CMR) can accurately identify myocardial scar. We performed a retrospective, single-center observational study to evaluate the association between the extent and distribution of left ventricular scar, quantified using LGE-CMR, and the burden of ventricular arrhythmias in patients with coronary artery disease and ICDs. METHODS AND RESULTS: All patients included (2006 to 2009) had undergone LGE-CMR before ICD implantation. Scar (defined as myocardium with a signal intensity ≥50% of the maximum in scar tissue) was characterized in terms of percent scar, scar surface area, and number of transmural left ventricular scar segments. The end point was appropriate ICD therapy. Sixty-four patients (mean age, 66±11 years; male sex, 51) were included. During 19±10 months follow-up, appropriate ICD therapy occurred in 19 (30%) patients. In Cox regression analyses, both percent scar (hazard ratio per 10%, 1.75; 95% CI, 1.09 to 2.81; P=0.02) and number of transmural scar segments (hazard ratio per segment, 1.40; 95% CI, 1.15 to 1.70; P=0.001) were significantly associated with the occurrence of appropriate ICD therapy. CONCLUSIONS: In this pilot study, the extent of myocardial scar characterized by LGE-CMR was significantly associated with the occurrence of spontaneous ventricular arrhythmias. We hypothesize that scar quantification by LGE-CMR may prove a valuable risk stratification tool for the occurrence of ventricular arrhythmias, which may have implications for patient selection for ICD therapy.