Subsequent Health Care Utilization Research Articles

Comparing mortality and healthcare utilization in the year following a paramedic-attended non-fatal overdose among people who were and were not transported to hospital: A prospective cohort study using linked administrative health data

BackgroundAs the overdose emergency continues in British Columbia (BC), paramedic-attended overdoses are increasing, as is the proportion of people not transported to hospital following an overdose. This study investigated risk of death and subsequent healthcare utilization for people who were and were not transported to hospital after a paramedic-attended non-fatal overdose. MethodsUsing a linked administrative health data set which includes all overdoses that come into contact with health services in BC, we conducted a prospective cohort study of people who experienced a paramedic-attended non-fatal overdose between 2015 and 2016. People were followed for 365 days after the index event. The primary outcomes assessed were all-cause mortality and overdose-related death. Additionally, we examined healthcare utilization after the index event. ResultsIn this study, 8659 (84%) people were transported and 1644 (16%) were not transported to hospital at the index overdose event. There were 279 overdose deaths (2.7% of people, 59.4% of deaths) during follow-up. There was no significant difference in risk of overdose-related death, though people not transported had higher odds of a subsequent non-fatal overdose event captured in emergency department and outpatient records within 90 days. People transported to hospital had higher odds of using hospital and outpatient services for any reason within 365 days. ConclusionsTransport to hospital after a non-fatal overdose is an opportunity to provide care for underlying and chronic conditions. There is a need to better understand factors that contribute to non-transport, particularly among people aged 20–59 and people without chronic conditions.

VA Antibiotic Stewardship Intervention to Improve Outpatient Antibiotic Use for ARIs: A Cost-Effectiveness Analysis

Background: The Core Elements of Outpatient Antibiotic Stewardship provide a framework to improve antibiotic use, but cost-effectiveness data on interventions to improve antibiotic use are limited. Beginning in September 2017, an antibiotic stewardship intervention was launched in within 10 outpatient Veterans Healthcare Administration clinics. The intervention was based on the Core Elements and used an academic detailing (AD) and an audit and feedback (AF) approach to encourage appropriate use of antibiotics. The objective of this analysis was to evaluate the cost-effectiveness of the intervention among patients with uncomplicated acute respiratory tract infections (ARI). Methods: We developed an economic simulation model from the VA’s perspective for patients presenting for an index outpatient clinic visit with an ARI (Fig. 1). Effectiveness was measured as quality-adjusted life-years (QALYs). Cost and utility parameters for antibiotic treatment, adverse drug reactions (ADRs), and healthcare utilization were obtained from the published literature. Probability parameters for antibiotic treatment, appropriateness of treatment, antibiotic ADRs, hospitalization, and return ARI visits were estimated using VA Corporate Data Warehouse data from a total of 22,137 patients in the 10 clinics during 2014–2019 before and after the intervention. Detailed cost data on the development of the AD and AF materials and electronically captured time and effort for the National AD Service activities by specific providers from a national ARI campaign were used as a proxy for the cost estimate of similar activities conducted in this intervention. We performed 1-way and probabilistic sensitivity analyses (PSAs) using 10,000 second-order Monte Carlo simulations on costs and utility values using their means and standard deviations. Results: The proportion of uncomplicated ARI visits with antibiotics prescribed (59% vs 40%) was lower and appropriate treatment was higher (24% vs 32%) after the intervention. The intervention was estimated to cost $110,846 (2018 USD) over a 2-year period. Compared to no intervention, the intervention had lower mean costs ($880 vs $517) and higher mean QALYs (0.837 vs 0.863) per patient because of reduced inappropriate treatment, ADRs, and subsequent healthcare utilization, including hospitalization. In threshold analyses, the antibiotic stewardship strategy was no longer dominant if intervention cost was >$64,415,000 or the number of patients cared for was <3,672. In the PSA, the antibiotic stewardship intervention was dominant in 100% of the 10,000 Monte Carlo iterations (Fig. 2). Conclusions: In every scenario, the VA outpatient AD and AF antibiotic stewardship intervention was a dominant strategy compared to no intervention.Funding: NoneDisclosures: None

Treatment Outcomes and Subsequent Healthcare Utilization Among Patients With Injection Drug Use–Associated Endocarditis

Background: Addiction medicine consultation and medication-assisted treatment (MAT) have been promoted as a way to improve outcomes for patients hospitalized with injection drug use–associated endocarditis (IDU-IE). However, IDU-IE outcomes have not been evaluated in settings where these services are commonplace. Objective: In this study, we evaluated IDU-IE outcomes in a setting where involvement of addiction medicine consultants and use of MAT is well integrated into patient care. Methods: Medical records of patients hospitalized with a diagnosis of bacteremia or infective endocarditis (IE) between October 1, 2015, and December 31, 2017, at a safety-net hospital in Boston were screened for evidence of active injection drug use (IDU) within 6 months of hospitalization (as documented by providers or as supported by urine toxicology assays) for suspected or definite IE using modified Duke criteria. Patients without active IDU or IE were excluded, as were those with a diagnosis of IDU-IE over the 6 months prior to the index hospitalization. Demographic parameters, receipt of antibiotics and MAT, other clinical information, and details of rehospitalizations were recorded. Analyses of descriptive statistics were performed. Results: Of 567 subjects screened for inclusion, 47 patients met inclusion criteria. All had opiate use disorder (OUD); 41 patients (87.2%) had polysubstance abuse. Addiction medicine consultation was completed for 41 patients (87.2%). Of the 47 subjects, 23 patients (54.8%) received MAT (methadone or buprenorphine/naloxone) over their entire hospitalization, and 31 patients (73.8%) received MAT for >75% of the index admission. Moreover, 43 patients (91.5%) survived to discharge, of whom 28 (59.6%) completed antibiotic therapy. Relapsed IDU was observed in 33 patients (76.7%). Relapsed IDU trended toward significance among undomiciled patients (OR, 4.07,; 95% CI, 0.93–17.85; P = .06). Also, 24 patients (55.8%) were rehospitalized within 1 year due to infectious complications of IDU; undomiciled patients were readmitted more frequently (OR, 20.45; 95% CI, 1.09–383.99; P = .04). Completion of IDU-IE antibiotic therapy, relapse of IDU, and rehospitalization were not associated with prior AMA discharges, duration or variety of IDU, receipt of MAT during the index admission, or addiction medicine consultation. The rate of readmission due to an infectious complication of IDU within 1 year was unrelated to the proportion of hospital days where MAT was prescribed. Conclusions: In settings with high rates of addiction medicine consultation and in-hospital MAT administration, inpatient interventions targeting OUD may not necessarily be protective against morbidity and rehospitalization. Focusing on housing instability and outpatient continuation of MAT may be beneficial.Funding: NoneDisclosures: None

Genetic Variant Reinterpretation: Economic and Population Health Management Challenges.

Genetic Variant Reinterpretation: Economic and Population Health Management Challenges.

The Effect of Including Benchmark Prevalence Data of Common Imaging Findings in Spine Image Reports on Health Care Utilization Among Adults Undergoing Spine Imaging

Lumbar spine imaging frequently reveals findings that may seem alarming but are likely unrelated to pain. Prior work has suggested that inserting data on the prevalence of imaging findings among asymptomatic individuals into spine imaging reports may reduce unnecessary subsequent interventions. To evaluate the impact of including benchmark prevalence data in routine spinal imaging reports on subsequent spine-related health care utilization and opioid prescriptions. This stepped-wedge, pragmatic randomized clinical trial included 250 401 adult participants receiving care from 98 primary care clinics at 4 large health systems in the United States. Participants had imaging of their backs between October 2013 and September 2016 without having had spine imaging in the prior year. Data analysis was conducted from November 2018 to October 2019. Either standard lumbar spine imaging reports (control group) or reports containing age-appropriate prevalence data for common imaging findings in individuals without back pain (intervention group). Health care utilization was measured in spine-related relative value units (RVUs) within 365 days of index imaging. The number of subsequent opioid prescriptions written by a primary care clinician was a secondary outcome, and prespecified subgroup analyses examined results by imaging modality. We enrolled 250 401 participants (of whom 238 886 [95.4%] met eligibility for this analysis, with 137 373 [57.5%] women and 105 497 [44.2%] aged >60 years) from 3278 primary care clinicians. A total of 117 455 patients (49.2%) were randomized to the control group, and 121 431 patients (50.8%) were randomized to the intervention group. There was no significant difference in cumulative spine-related RVUs comparing intervention and control conditions through 365 days. The adjusted median (interquartile range) RVU for the control group was 3.56 (2.71-5.12) compared with 3.53 (2.68-5.08) for the intervention group (difference, -0.7%; 95% CI, -2.9% to 1.5%; P = .54). Rates of subsequent RVUs did not differ between groups by specific clinical findings in the report but did differ by type of index imaging (eg, computed tomography: difference, -29.3%; 95% CI, -42.1% to -13.5%; magnetic resonance imaging: difference, -3.4%; 95% CI, -8.3% to 1.8%). We observed a small but significant decrease in the likelihood of opioid prescribing from a study clinician within 1 year of the intervention (odds ratio, 0.95; 95% CI, 0.91 to 1.00; P = .04). In this study, inserting benchmark prevalence information in lumbar spine imaging reports did not decrease subsequent spine-related RVUs but did reduce subsequent opioid prescriptions. The intervention text is simple, inexpensive, and easily implemented. ClinicalTrials.gov Identifier: NCT02015455.

Increased Costs and Health Care Utilization Associated with Delay in Diagnosis of ACL Injuries

Objectives:Anterior cruciate ligament (ACL) injuries can be difficult to diagnose, especially among providers with limited musculoskeletal training. Early and accurate diagnosis of ACL injuries is important for initiating proper treatment and limiting inefficient or costly healthcare utilization. We hypothesized that a delayed diagnosis in patients sustaining ACL injuries would result in treatment that is more costly and utilizes more healthcare resource compared to patients diagnosed without delay.Methods:A large national private insurance database (Truven MarketScan) was utilized to identify all patients aged 10-63 years-old diagnosed with an ACL tear from 2011 through 2015. Patients with a possible delay in diagnosis were determined via a database search for the most commonly assigned alternative knee-related diagnosis codes found in patients who were later diagnosed with an ACL tear. Patients were stratified into three groups according to time from non-specific knee diagnosis code to ACL diagnosis: 1) no delay (0 days), short delay (1-30 days) and long delay (>30 days). Subsequent cost and healthcare utilization data including physical therapy, medication use, and number of clinic visits were collected for a period of one-year after initial presentation with an ACL tear or knee-related diagnosis.Results:A search from 2011 through 2015 identified 87,435 patients meeting inclusion criteria that were diagnosed with an ACL injury. There were 24.1% of patients diagnosed at initial presentation whereas 43.5% were diagnosed between one day and 30 days. There were 38.2% of patients that were treated without surgery and had the lowest total cost regardless of delay in diagnosis (Table 1). ACL reconstruction combined with other knee procedures had the highest total cost, particularly with a longer delay in diagnosis. Each day of delay in diagnosis was three times more costly in the non-surgical (additional $27.06 cost per day of delay) versus the surgical groups ($8.09 cost per day of delay).Conclusions:A timely diagnosis of ACL injuries is associated with a lower cost of treatment and less visits to healthcare providers. Non-operative injuries may benefit the most from a timely diagnosis given the high additional cost per day of diagnosis delay. In order to further decrease the burden on the healthcare system, future studies should assess methods than can decrease delays in diagnosis while measuring cost and utilization differences as well as outcomes after ACL tears.

Long-term healthcare utilisation following child sex abuse: A follow-up study utilising five years of medical data

BackgroundLimited attention has been paid to the long-term physical health consequences experienced by CSA survivors. Research has found that CSA is often associated with increased health burden in adulthood. However, research in this area is plagued by a range of methodological difficulties, rendering it difficult to draw conclusions regarding the health status of CSA survivors. ObjectiveThis research sought to investigate associations between child sexual abuse (CSA) victimisation and rates of subsequent healthcare utilisation. MethodThe forensic medical records of 2759 cases of CSA between 1964 and 1995 were linked to commonwealth Australian medical data recorded between 2010 and 2015. Differences in rates of health service utilisation during this five-year period were compared between CSA victims and a age and gender matched comparison cohort without known abuse history. FindingsCSA was associated with increased levels of attendance at health professionals (OR = 1.51, p < .001) for medical care and for consultations related to dental (OR = 1.28, p < .001) and chronic disease (OR = 1.23, p <.001). CSA was associated with lower mean rates of attendance at gynaecologists. Gender and age at abuse significantly influenced the relationship between CSA and frequency of health service utilisation. ConclusionsExperiencing CSA was associated with increases in rates of contact with health professionals. Abused females and victims abused after 12 years of age demonstrated the greatest elevations in rates of service utilisation. These findings have significance for both researchers in this field, and medical practitioners providing primary care.

Queensland Cardiovascular Data Linkage (QCard): A population-based cohort study

Queensland is Australia's hotspot for cardiovascular disease (CVD). Critically, beyond modifiable lifestyle risk factors; socio-demographic differences and environmental factors account for significant variations in healthcare use and outcomes among cardiac patients across the country. To better understand the impacts of these factors on the health of cardiac patients, there is a need for a comprehensive and robust longitudinal cohort study that can unpack the underlying dynamics. This paper describes the protocol for the Queensland Cardiovascular Linkage (QCard) Study. The QCard is a longitudinal linkage cohort study of cardiac patients who were first hospitalised with any cardiac condition in 2010, with follow up hospitalisations until December 2015. The primary aim of the QCard is to identify and characterise the nature and impact of socio-demographic inequality among those presenting for the first time with the most common form of CVD in Australia (heart disease) in Queensland from 2010 with minimum 5-years follow-up of subsequent healthcare utilisation and outcomes. A secondary aim is to undertake an exploration of the impact of environmental and specific health service factors on healthcare use and survival time in the same QCard cohort. Administrative public and private hospital inpatient, outpatient and emergency department data for all of Queensland will be linked with individual primary care data and pharmaceutical data. These data will also be linked to regional socio-demographic data and environmental data, as well as data that describes the features of each hospital in the region. The findings from the study will provide critical information for cardiac patients, clinicians and health policymakers. Such information ranges from identifying most vulnerable cardiac patients who may require targeted needs to providing estimates for cost-effective ways of evaluating healthcare interventions that seek to improve the health of cardiac patients.

Abstract P2-08-13: Genetic testing for breast cancer and associated downstream utilization and costs

Abstract Background: Given the clinical utility for diagnosis and treatment of hereditary breast cancer, the use of genetic testing is growing. However, few data exist on the association between genetic testing and subsequent clinical outcomes, healthcare utilization, and cost. Objective: To examine medical utilization, cost, and survival in individuals with breast cancer who were tested for genes with high evidence for predicting breast cancer risk, as compared to those who were untested. Methods: A retrospective observational study was conducted in commercial claims. Adult women with non-metastatic breast cancer during 2013-2017 were identified and divided into two groups: (1) a genetic testing group (GTG) that was tested during January 2014-June 2017 for high risk genes for hereditary breast cancer (BRCA, PTEN) and (2) a no genetic testing group (NGTG) that did not test for any genes or gene expression for hereditary breast cancer. Members in the NGTG were matched 1:1 to the GTG by propensity scores using age, region, baseline clinical variables (other cancer types, breast imaging, breast biopsy, Deyo-Charlson Comorbidity Index), and baseline all-cause and cancer-related costs. Utilization and cost outcomes were assessed during 1 year post-genetic testing. We also estimated survival and all-cause healthcare costs during full follow-up, in the main sample and in a “3-year” sample of women who took their first genetic test at least 3 years prior to study end. Variables remained unbalanced after matching were adjusted in regression modeling. Results: The study identified 12,236 patients in the GTG and 53,125 patients in the NGTG. All patient covariates, except baseline cancer-related costs, were balanced after 1:1 propensity score matching (n=10,887 matched patients in each group). In the GTG and NGTG, 18% and 1%, respectively, had follow-up genetic testing (p&amp;lt;.001). The GTG, compared to the matched NGTG, was more likely to have breast imaging (adjusted odds ratio [aOR] 1.43; 95% confidence interval [CI] 1.35-1.51), breast biopsy (aOR 2.24; 95% CI 2.02-2.49), breast surgery (aOR 2.76; 95% CI 2.62-2.91), and other cancer treatments during 1 year of follow-up (Table; all p&amp;lt;0.001, except mammogram p-value=0.002). All-cause healthcare cost was $27,131 higher among GTG compared to NGTG during 1 year of follow-up ($78,733 versus $51,602; p&amp;lt;0.001) and remained high at 3 years ($55,479 versus $38,037 per-member-per-year; p&amp;lt; 0.001). Genetic testing for breast cancer was not associated with difference in survival (overall adjusted hazard ratio [aHR] 1.17, 95% CI 0.82-1.67; 3-year aHR 0.77, 95% CI 0.23-2.63). Conclusions: Increased treatment and healthcare costs are likely to follow genetic testing and should be discussed in the shared decision-making process. Survival analyses were inconclusive due to low rate of mortality. Citation Format: Shia T Kent, Xiaoxue Chen, David Debono, Karen Lewis, Abiy Agiro. Genetic testing for breast cancer and associated downstream utilization and costs [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P2-08-13.

Impact of Atypical Antipsychotics as Adjunctive Therapy on Psychiatric Cost and Utilization in Patients with Major Depressive Disorder.

IntroductionPatients with major depressive disorder (MDD) incur high costs, despite established treatment options. Adding an atypical antipsychotic (AAP) to antidepressant therapy has shown to reduce depressive symptoms in MDD, but it remains unclear with which adjunctive AAP to initiate. As economic burden is one factor that can influence treatment selection, this study’s objective was to evaluate the impact of adjunctive AAP choice on psychiatric costs and healthcare utilization in MDD.Materials and MethodsThis retrospective cohort study analyzed de-identified data from: (1) IBM® MarketScan® Commercial (C), Medicare Supplemental (MS), and MarketScan Multi-State Medicaid (M) Databases, and (2) Optum® Clinformatics® Datamart. Adult MDD patients were included if they had: initiated adjunctive AAPs during study identification period (7/1/15-9/30/16 MarketScan C/MS, and Optum; 7/1/15-6/30/16 MarketScan M), and ≥12 months of continuous enrollment before (baseline) and after (follow-up) first treatment date. Models included generalized linear models (GLMs) for psychiatric costs (total inpatient and outpatient services, excluding outpatient pharmacy costs), and a two-part model (logistic regression for psychiatric hospitalizations, GLM for psychiatric hospitalization costs among hospitalized patients); models were adjusted for baseline characteristics.ResultsThe final study sample consisted of 10,325 patients (7657 aripiprazole, 1219 brexpiprazole, 827 lurasidone, 622 quetiapine). Using brexpiprazole as reference, lurasidone and quetiapine users had $1662 and $3894 higher psychiatric costs, respectively. Psychiatric costs were not statistically significantly different between aripiprazole and brexpiprazole (p>0.05). Quetiapine users had $15,159 (p<0.001) higher psychiatric hospitalization costs among those hospitalized, and higher odds of psychiatric hospitalization [2.11 (1.46–3.04); p<0.001] compared to brexpiprazole users. No statistically significant differences observed in psychiatric hospitalization risk comparing aripiprazole and lurasidone with brexpiprazole (p>0.05).ConclusionIn MDD, brexpiprazole users had significantly lower psychiatric costs than lurasidone and quetiapine users, and significantly lower psychiatric hospitalization risk than quetiapine users. Adjunctive AAP choice may impact subsequent healthcare costs and utilization in MDD.

Impact of Gastrointestinal Surgical Site Wound Complications on Long-term Healthcare Utilization

ImportanceWound complication following gastrointestinal surgery substantially impacts the quality and costs of surgical care. The impact of wound complication on subsequent long-term healthcare utilization has not been fully studied. ObjectiveWe assessed the impact of surgical wound complication on inpatient and outpatient healthcare utilization in the 2 years after gastrointestinal (GI) surgery. DesignAn observational retrospective cohort study was conducted on Veterans Affairs health system patients who underwent an inpatient GI surgical procedure, were assessed by the Veterans Affairs Surgical Quality Improvement Program (VASQIP), and were discharged alive from Veterans Affairs (VA) hospitals between October 1, 2007 and September 30, 2014. SettingPopulation-based ParticipantsA total of 64,351 patients underwent a GI surgical procedure in the VA system between 2007 and 2014. The cohort was 93.5% male, with a median age of 63.0 years (interquartile range (IQR) 57.0–70.0). A total of 7880 patients (12.2%) had at least one reported wound complication, 5460 of which had their postoperative wound complication classified by a VASQIP nurse. ExposureVASQIP-assessed or ICD-9-coded wound complication in the 30 days after surgery Main Outcome MeasurementsInpatient visits, total inpatient days, outpatient visits, and emergency department visits, and operative interventions up to 2 years after discharge from index admission ResultsPatients with a postoperative wound complication had greater inpatient healthcare utilization compared with no-wound complication for up to 2 years after surgery: inpatient admissions (mean number 3.5 vs. 2.8; P < .001), inpatient bed days (mean 41.0 vs. 25.0; P < .001). Patients with a postoperative wound complication also had greater 2-year outpatient utilization than the no-wound complication cohort: outpatient visits (mean number 92.7 vs. 75.9; P < .001) and emergency department visits (mean 3.5 vs. 2.7; P < .001). The same relationship held for wound-related parameters; inpatient admissions (2.2 vs. 0.4; P < .001); inpatient bed days (21.4 vs. 3.7; P < .001); and outpatient visits (56.2 vs. 9.7; P < .001). A greater proportion of patients in the wound complication cohort had an operative intervention for all time intervals examined (P < .001). ConclusionsSurgical wound complications impact healthcare utilization patterns for up to 2 years after the index procedure including hospital readmissions and operative interventions; efforts to reduce postoperative wound complications will have substantial effects on patient outcomes and healthcare expenditures well beyond the 30-day postoperative period.

More than expected? Assessing departures from additivity in health care utilization after mild traumatic brain injury in military personnel with pre-existing mental health conditions

Objective Prior research indicates that there is an additive association between traumatic brain injury and mental health diagnoses on health-care utilization. This assumed additivity has not been formally assessed. The objective of this study was to estimate additive and multiplicative interactions associated with mild traumatic brain injury (mTBI) and pre-existing health conditions. Method Active-duty military patient records over a nine-year period were sampled within four exposure groups (N = 4500 per group) defined jointly by incident mTBI and pre-existing mental health diagnoses. Outpatient and inpatient health encounters were compared between the four exposure groups using generalized linear models for count and proportion outcomes. Additive interactions were estimated using the interaction contrast ratio. Multiplicative interactions were estimated as a product term in the generalized linear models. Results The joint association of mTBI and pre-existing mental health diagnoses with health-care utilization, overall, was less than multiplicative and greater than additive. Patients with both exposures experienced more health-care utilization than expected under the assumed additivity (independence) of the two exposures. PTSD and anxiety diagnoses were the MH diagnoses associated with the largest interaction contrast values specific to total outpatient encounters. Conclusions Studies of the interaction of two diagnoses on subsequent health-care utilization should examine both additive and multiplicative interactions. The greater-than-additive findings in this study indicate that there may be synergy, for at least some patients, between mTBI injury and mental health that complicates the treatment course.

Evaluation of machine learning methodology for the prediction of healthcare resource utilization and healthcare costs in patients with critical limb ischemia-is preventive and personalized approach on the horizon?

Critical limb ischemia (CLI) is a severe stage of peripheral arterial disease and has a substantial disease and economic burden not only to patients and families, but also to the society and healthcare systems. We aim to develop a personalized prediction model that utilizes baseline patient characteristics prior to CLI diagnosis to predict subsequent 1-year all-cause hospitalizations and total annual healthcare cost, using a novel Bayesian machine learning platform, Reverse Engineering Forward Simulation™ (REFS™), to support a paradigm shift from reactive healthcare to Predictive Preventive and Personalized Medicine (PPPM)-driven healthcare. Patients ≥ 50years with CLI plus clinical activity for a 6-month pre-index and a 12-month post-index period or death during the post-index period were included in this retrospective cohort of the linked Optum-Humedica databases. REFS™ built an ensemble of 256 predictive models to identify predictors of all-cause hospitalizations and total annual all-cause healthcare costs during the 12-month post-index interval. The mean age of 3189 eligible patients was 71.9years. The most common CLI-related comorbidities were hypertension (79.5%), dyslipidemia (61.4%), coronary atherosclerosis and other heart disease (42.3%), and type 2 diabetes (39.2%). Post-index CLI-related healthcare utilization included inpatient services (14.6%) and ≥ 1 outpatient visits (32.1%). Median annual all-cause and CLI-related costs per patient were $30,514 and $2196, respectively. REFS™ identified diagnosis of skin and subcutaneous tissue infections, cellulitis and abscess, use of nonselective beta-blockers, other aftercare, and osteoarthritis as high confidence predictors of all-cause hospitalizations. The leading predictors for total all-cause costs included region of residence and comorbid health conditions including other diseases of kidney and ureters, blindness of vision defects, chronic ulcer of skin, and chronic ulcer of leg or foot. REFS™ identified baseline predictors of subsequent healthcare resource utilization and costs in CLI patients. Machine learning and model-based, data-driven medicine may complement physicians' evidence-based medical services. These findings also support the PPPM framework that a paradigm shift from post-diagnosis disease care to early management of comorbidities and targeted prevention is warranted to deliver a cost-effective medical services and desirable healthcare economy.

Multidimensional sleep health and subsequent health-care costs and utilization in older women.

Determine the association of poor multidimensional sleep health with health-care costs and utilization. We linked 1,459 community-dwelling women (mean age 83.6 years) participating in the Study of Osteoporotic Fractures Year 16 visit (2002-2004) with their Medicare claims. Five dimensions of sleep health (satisfaction, daytime sleepiness, timing, latency, and duration) were assessed by self-report. The number of impaired dimensions was expressed as a score (range 0-5). Total direct health-care costs and utilization were ascertained during the subsequent 36 months. Mean (SD) total health-care costs/year (2017 dollars) increased in a graded manner across the sleep health score ranging from $10,745 ($15,795) among women with no impairment to up to $15,332 ($22,810) in women with impairment in three to five dimensions (p = 0.01). After adjustment for age, race, and enrollment site, women with impairment in three to five dimensions vs. no impairment had greater mean total costs (cost ratio [CR] 1.34 [95% CI = 1.13 to 1.60]) and appeared to be at higher risk of hospitalization (odds ratio (OR) 1.31 [95% CI = 0.96 to 1.81]). After further accounting for number of medical conditions, functional limitations, and depressive symptoms, impairment in three to five sleep health dimensions was not associated with total costs (CR 1.02 [95% CI = 0.86 to 1.22]) or hospitalization (OR 0.91 [95% CI = 0.65 to 1.28]). Poor multidimensional sleep health was not related to outpatient costs or risk of skilled nursing facility stay. Older women with poor sleep health have higher subsequent total health-care costs largely attributable to their greater burden of medical conditions, functional limitations, and depressive symptoms.

Advance Care Planning Claims and Health Care Utilization Among Seriously Ill Patients Near the End of Life

Although advance care planning is known to increase patient and caregiver satisfaction, its association with health care utilization is not well understood. To examine the association between billed advance care planning encounters and subsequent health care utilization among seriously ill patients. This retrospective cohort study conducted from October 1, 2015, to May 31, 2018, used a national commercial insurance claims database to retrieve data from 18 484 Medicare Advantage members 65 years or older who had a claim that contained a serious illness diagnosis. A claim that contained an advance care planning billing code between October 1, 2016, and November 30, 2017. Receipt of intensive therapies, hospitalization, emergency department use, hospice use, costs, and death during the 6-month follow-up period. The final study sample included 18 484 seriously ill patients (mean [SD] age, 79.7 [7.9] years; 10 033 [54.3%] female), 864 (4.7%) of whom had a billed advanced care planning encounter between October 1, 2016, and November 30, 2017. In analyses adjusted for patient characteristics and a propensity score for advance care planning, the presence of a billed advance care planning encounter was associated with a higher likelihood of hospice enrollment (incidence rate ratio [IRR], 2.52; 95% CI, 2.22-2.86) and mortality (hazard ratio, 2.27; 95% CI, 1.79-2.88) compared with no billed advance care planning encounter. Although patients with billed advance care planning encounters were also more likely to be hospitalized (IRR, 1.37; 95% CI, 1.26-1.49), including in the intensive care unit (IRR, 1.25; 95% CI, 1.08-1.45), they were less likely to receive any intensive therapies (IRR, 0.85; 95% CI, 0.78-0.92), such as chemotherapy (IRR, 0.65; 95% CI, 0.55-0.78). Similar results were observed in a propensity score-matched analysis (99% matched) and in a decedent analysis of patients who died during the 6-month follow-up period. Patients with billed advance care planning encounters were more likely than those without these encounters to receive hospice services and less likely to receive any intensive therapies, such as chemotherapy. However, they were also hospitalized more frequently than patients without billed advance care planning encounters. Although these findings were robust to multiple analytic methods, the results may be attributable to residual confounding because of a higher unmeasured severity of illness in the advance care planning group. Additional evidence appears to be needed to understand the effect of advance care planning on these outcomes.

Association of sleep-disordered breathing with total healthcare costs and utilization in older men: the Outcomes of Sleep Disorders in Older Men (MrOS Sleep) study.

To determine the associations of sleep-disordered breathing (SDB) with subsequent healthcare costs and utilization including inpatient and post-acute care facility stays among community-dwelling older men. Participants were 1,316 men (mean age 76.1 [SD = 5.7] years) in the Outcomes of Sleep Disorders in Older Men (MrOS sleep) study (from December 2003 to March 2005), who were enrolled in a Medicare Fee-For-Service plan. Primary SDB measures including apnea hypopnea index (AHI) and oxygen desaturation index (ODI) were collected using in-home level 2 polysomnography. Incident healthcare costs and utilization were determined from claims data in the subsequent 3-year period post-MrOS sleep visit. Five hundred and twenty-nine (40.2%) men had at least one hospitalization in the 3-year period. Compared with those without sleep apnea (AHI < 5/hour), men with moderate to severe sleep apnea (AHI ≥ 15/hour) had a higher odds of all-cause hospitalization (odds ratio [OR] adjusted for age and site 1.43, 95% confidence interval [CI]: 1.07-1.90). This association was slightly attenuated after further adjustment for traditional prognostic factors including education, body mass index, comorbid medical conditions, and health status (OR = 1.36; 95% CI: 1.01-1.83). Similar associations were observed for ODI. However, measures of SDB were not related to subsequent healthcare costs (total or outpatient) or odds of post-acute skilled nursing facility stay. Older men with SDB have an increased risk of hospitalization, not entirely explained by the greater prevalence of comorbid conditions, but not higher subsequent total healthcare costs. These findings indicate a need to evaluate the impact of SDB treatment on subsequent healthcare utilization.

Risk factors for healthcare utilization in patients with newly diagnosed advanced lung cancer.

111 Background: Advanced lung cancer (ALC) is a symptomatic disease often diagnosed in the context of hospitalization. The index hospitalization may be a window of opportunity to improve care delivery. We aimed to identify newly diagnosed ALC patients at highest risk for subsequent healthcare utilization. Methods: We identified patients in SEER-Medicare with: ALC (stage IIIB-IV SCLC or NSCLC), diagnosed 2007-13; continuous enrollment -6 months from diagnosis through death or 12/2014; and an index hospitalization within 7 days of ALC diagnosis. Our primary outcomes were 30-day re-hospitalization and emergency department (ED) use. We used a time-to-event model with multivariate regression to identify risk factors. Results: Of ALC patients in SEER-Medicare, 54% (n=28,976) had an index hospitalization. Of those who survived to discharge, by 6 months, 47% had been re-hospitalized, and 50% had enrolled in hospice. Only 37% of patients ever received systemic treatment for their cancer, and the median time to treatment was 46 days. The 30-day ED visit and readmission rates were 13% and 35%, respectively. Pre-cancer ED use or hospitalization, SCLC, and prolonged length of stay were associated with higher risk of 30-day utilization. Palliative care consultation and discharge to hospice were associated with substantially lower risk of 30-day readmission. Conclusions: Many newly diagnosed ALC patients experience an early return to the hospital. These patients may benefit from increased access to palliative and other supportive care during index hospitalization to prevent subsequent healthcare utilization. [Table: see text]

Deficits in the Palliative Care Process Measures in Patients with Advanced Pancreatic Cancer Undergoing Operative and Invasive Nonoperative Palliative Procedures.

Given survival measured in months, metrics, such as 30-day mortality, are poorly suited to measure the quality of palliative procedures for patients with advanced cancer. Nationally endorsed process measures associated with high-quality PC include code-status clarification, goals-of-care discussions, palliative-care referral, and hospice assessment. The impact of the performance of these process measures on subsequent healthcare utilization is unknown. Administrative data and manual review were used to identify hospital admissions with performance of palliative procedures for advanced pancreatic cancer at two tertiary care hospitals from 2011 to 2016. Natural language processing, a form of computer-assisted abstraction, identified process measures in associated free-text notes. Healthcare utilization was compared using a Cox proportional hazard model. We identified 823 hospital admissions with performance of a palliative procedure. PC process measures were identified in 68% of admissions. Patients with documented process measures were older (66 vs. 63; p = 0.04) and had a longer length of stay (9 vs. 6days; p < 0.001). In multivariate analysis, patients treated by surgeons were less likely to have PC process measures performed (odds ratio 0.19; 95% confidence interval 0.10-0.37). Performance of PC process measures was associated with decreased healthcare utilization in a Cox proportional hazard model. PC process measures were not performed in almost one-third of hospital admissions for palliative procedures in patients with advanced pancreatic cancer. Performance of established high-quality process measures for seriously ill patients undergoing palliative procedures may help patients to avoid burdensome, high-intensity care at the end-of-life.

An Intervention for Person-Centered Support in Irritable Bowel Syndrome: Development and Pilot Study.

Irritable bowel syndrome is a common and often chronic functional bowel disorder that can cause severe disruption of daily functioning in those affected, with subsequent high healthcare utilization and work absenteeism. Nurses represent an underutilized group in the current management of irritable bowel syndrome. The aim of this study was to systematically develop a person-centered support intervention in irritable bowel syndrome and evaluate this in a pilot study. The development followed the revised framework for complex interventions from the Medical Research Council and involved literature reviews and multiprofessional expert groups. The intervention was then tested in a pilot study including 17 patients and evaluated through validated questionnaires measuring irritable bowel syndrome symptom severity, gastrointestinal-specific anxiety, and self-efficacy as well as through interviews. There was a significant improvement in irritable bowel syndrome symptom severity between baseline and follow-up, but not for self-efficacy or gastrointestinal-specific anxiety. The patients' perceptions of participating in the intervention were positive and induced a learning process; they were able to form a supportive relationship with the nurse and their ability to self-manage improved. The promising results from this small pilot study in terms of feasibility, potential efficacy, and the patients' positive feedback make this intervention a suitable candidate for a larger controlled trial.

A Predictive Model for Identification of Children at Risk of Subsequent High-Frequency Utilization of the Emergency Department for Asthma.

Asthma is the most common chronic condition among children with high-frequency emergency department (ED) utilization. Previous research has shown in outpatients seen for asthma that acute care visits predict subsequent health care utilization. Among ED patients, however, the optimal method of predicting subsequent ED utilization remains to be described. The goal of this study was to create a predictive model to identify children in the ED who are at risk of subsequent high-frequency utilization of the ED for asthma. We used 3 years of data, 2013-2015, drawn from the electronic health records at a tertiary care, urban, children's hospital that is a high-volume center for asthma care. Data were split into a derivation (50%) and validation/test (50%) set, and 3 models were created for testing: (1) all index patients; (2) removing patients with complex chronic conditions; and (3) subset of patients with in-network care on whom more clinical data were available. Each multivariable model was then tested in the validation set, and its performance evaluated by predicting error rate, calculation of a receiver operating characteristic (ROC) curve, and identification of the optimal cutpoint to maximize sensitivity and specificity. There were 5535 patients with index ED visits, of whom 2767 were in the derivation set and 2768 in the validation set. Of the 5535 patients, 125 patients (2.3%) had 4 or more visits for asthma in the outcome year. Significant predictors in models 1 and 2 were age and number of prior ED visits for asthma. For model 3 (additional clinical information available), the predictors were number of prior ED visits for asthma, number of primary care visits, and not having a controller medication. Areas under the ROC curve were 0.77 for model 1, 0.80 for model 2, and 0.77 for model 3. Administrative data available at the time of ED triage can predict subsequent high utilization of the ED, with areas under the ROC curve of 0.77 to 0.80. The addition of clinical variables did not improve the model performance. These models provide useful tools for researchers interested in examining intervention efficacy by predicted risk group.