Advertisement

Determinants of Admission to Inpatient Rehabilitation Among Acute Care Survivors of Hypoxic-Ischemic Brain Injury: A Prospective Population-Wide Cohort Study

Published:January 29, 2016DOI:https://doi.org/10.1016/j.apmr.2016.01.007

      Abstract

      Objective

      To investigate demographic and acute care clinical determinants of admission to inpatient rehabilitation (IR) among patients with hypoxic-ischemic brain injury (HIBI) who survive the initial acute care episode.

      Design

      Population-wide prospective cohort study using Canadian Institutes for Health Information administrative health data from Ontario, Canada. All patients who survived their HIBI acute care episode during the study period remained eligible for the outcome, admission to IR, for 1 year postacute care discharge.

      Setting

      Inpatient rehabilitation.

      Participants

      We included all patients with HIBI using International Classification of Diseases, Tenth Revision, Canadian Enhancement codes recorded at acute care admission who were ≥20 years old (N=599) and discharged from acute care between the 2002 and 2010 fiscal years, inclusive. Six patients were excluded from analyses because of missing data.

      Interventions

      Not applicable.

      Main Outcome Measure

      Admission to IR.

      Results

      Of HIBI survivors admitted to IR within 1 year of acute care discharge (n=169), most (56.2%) had an IR admitting diagnosis indicating anoxic brain damage. Younger age, being a man, lower comorbidity burden, longer length of stay of preceding acute care episode, and shorter duration in special care were most predictive of admission to IR in multivariable regression models. Women had an almost 2-fold lower incidence of admission to IR (risk ratio, .62; 95% confidence interval, .46–.84).

      Conclusions

      Older age, higher comorbidity burden, and shorter lengths of stay and delayed discharge from acute care are associated with lower incidence of IR admission for patients with HIBI. That women are almost 2-fold less likely to receive rehabilitation requires further investigation.

      Keywords

      List of abbreviations:

      ABI (acquired brain injury), ALC (alternate level of care), HIBI (hypoxic-ischemic brain injury), ICD-10 (International Classification of Diseases, 10th Revision), IR (inpatient rehabilitation), LOS (length of stay), NTBI (nontraumatic brain injury), TBI (traumatic brain injury)
      An audio podcast accompanies this article.
      Acquired brain injury (ABI) arising from either traumatic brain injury (TBI) or nontraumatic brain injury (NTBI) is a leading cause of death and disability worldwide.
      • Toronto ABI Network
      • Basso A.
      • Previgliano I.
      • Servadei F.
      Neurological disorders: a public health approach. Traumatic brain injuries.
      TBI is defined as damage to the brain by an external force, whereas its counterpart, NTBI, is more heterogeneously defined, capturing all other ABI etiologies (eg, stroke, nontraumatic hemorrhage, tumor, infectious diseases, hypoxic injuries, metabolic disorders, toxic exposure). Given this heterogeneity, evaluation of health services outcomes by distinct NTBI subgroup is warranted. Diverse injury mechanisms can lead to important distinctions in etiology and prognosis, the investigation of which can inform tailoring of more specific and efficient NTBI rehabilitation guidelines.
      Adult-onset hypoxic-ischemic brain injury (HIBI) is a sparsely studied NTBI caused by a deficient supply of oxygen to the brain.
      • Ropper A.H.
      • Samuels M.A.
      The acquired metabolic disorders of the nervous system.
      • Arciniegas D.B.
      Hypoxic-ischemic brain injury: addressing the disconnect between pathophysiology and public policy.
      Recent data from Ontario, Canada, demonstrates that approximately 10% of all NTBI cases in the emergency department and 15% of all NTBI cases in acute care were caused by HIBI.

      Colantonio A, Chan V, Zagorski B, Parsons D, Vander Laan R. Ontario Acquired Brain Injury (ABI) Dataset Project Phase III: highlights: number of episodes of care and causes of brain injury. Available at: http://www.abiresearch.utoronto.ca/ABI%20Dataset%20+%20LHIN%20USE%20THIS/LHIN%20Factsheets%20Demographics%20-%20June%206.%202013%20-%20REVISED%20WITH%20NEW%20LINK.pdf. Accessed June 17, 2015.

      The poor outcomes experienced by patients with HIBI are well-documented and include cognitive impairments, neurologic abnormalities, motor deficits, and psychosocial and psychological difficulties.
      • Wilson B.A.
      Cognitive functioning of adult survivors of cerebral hypoxia.
      • Lu-Emerson C.
      • Khot S.
      Neurological sequelae of hypoxic-ischemic brain injury.
      • Khot S.
      • Tirschwell D.L.
      Long-term neurological complications after hypoxic-ischemic encephalopathy.
      • Wilson M.
      • Staniforth A.
      • Till R.
      • das Nair R.
      • Vesey P.
      The psychosocial outcomes of anoxic brain injury following cardiac arrest.
      • Middelkamp W.
      • Moulaert V.
      • Verbunt J.A.
      • van Heugten C.M.
      • Bakx W.G.
      • Wade D.T.
      Life after survival: long-term daily life functioning and quality of life of patients with hypoxic injury as a result of cardiac arrest.
      Consequently, providing appropriate rehabilitation for this group is of the utmost importance. Although the evidence supports the effectiveness of postacute care rehabilitation among patients with HIBI, it suggests that its efficiency could be increased. For example, HIBI sufferers demonstrate lower cognitive and motor outcomes at discharge from rehabilitation relative to patients with TBI matched on demographic and clinical characteristics
      • Cullen N.K.
      • Park Y.
      • Bayley M.T.
      Functional recovery following traumatic vs. non-traumatic brain injury: a case-controlled study.
      • Cullen N.
      • Crescini C.
      • Bailey M.
      Rehabilitation outcomes after anoxic brain injury: a case controlled comparison with traumatic brain injury.
      • Cullen N.K.
      • Weisz K.
      Cognitive correlates with functional outcomes after anoxic brain injury: a case-controlled comparison with traumatic brain injury.
      and are less likely to be discharged home after rehabilitation.
      • Smania N.
      • Avesani R.
      • Roncari L.
      • et al.
      Factors predicting functional and cognitive recovery following severe traumatic, anoxic, and cerebrovascular brain damage.
      Additionally, access issues may plague admission to rehabilitation in this group because they have been observed less likely to be referred, and if referred, to wait longer for admission to rehabilitation relative to other brain injury populations.
      • Smania N.
      • Avesani R.
      • Roncari L.
      • et al.
      Factors predicting functional and cognitive recovery following severe traumatic, anoxic, and cerebrovascular brain damage.
      • Fitzgerald A.
      • Aditya H.
      • Prior A.
      • McNeill E.
      • Pentland B.
      Anoxic brain injury: clinical patterns and functional outcomes: a study of 93 cases.
      • Fertl E.
      • Vass K.
      • Sterz F.
      • Gabriel H.
      • Auf E.
      Neurological rehabilitation of severely disabled cardiac arrest survivors. Part 1. Course of post-acute inpatient treatment.
      • Chan V.
      • Zagorski B.
      • Parsons D.
      • Colantonio A.
      Older adults with acquired brain injury: a population based study.
      It is plausible that access to supplemental insurance is contributing to this discrepancy because, relative to TBI, many patients with HIBI would be less likely to receive benefits associated with injuries such as those sustained from motor vehicle collisions.
      • Chan V.
      • Zagorski B.
      • Parsons D.
      • Colantonio A.
      Older adults with acquired brain injury: a population based study.
      • Kraus J.F.
      • Chu L.D.
      Epidemiology.
      Although a number of studies have examined referral to rehabilitation among the TBI population,
      • Jourdan C.
      • Bayen E.
      • Bosserelle V.
      • et al.
      Referral to rehabilitation after severe traumatic brain injury: results from the PariS-TBI Study.
      • Foster M.
      • Tilse C.
      • Fleming J.
      Referral to rehabilitation following traumatic brain injury: practitioners and the process of decision-making.
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      to date, no population-wide study has focused on survivors of HIBI. This paucity of information limits our understanding of the apparent reduced effectiveness of inpatient rehabilitation (IR) relative to TBI. Using a cohort design, we identified patients with HIBI who survived their acute care episode in Ontario, Canada, from administrative health data and describe determinants of their admission to IR. Extending on prior evidence indicating that determinants of discharge destination from acute care differ across types of ABI,
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      the objective of this study was to investigate predictors of IR specific to survivors of HIBI, focusing on characteristics of the acute care episode, prior health care utilization, and sociodemographic factors. Our findings will potentially inform rehabilitation program planning for HIBI survivors and help identify gaps and inequities in access to IR among this distinct NTBI population.

      Methods

       Study design and data sources

      All data used herein were provided through the Ontario Cancer Data Linkage Program, an initiative of the Ontario Institute for Cancer Research/Cancer Care Ontario Health Services Research Program, whereby risk-reduced coded data from the Institute for Clinical Evaluative Sciences Data Repository managed by the Institute for Clinical Evaluative Sciences are provided directly to researchers with the protection of a comprehensive data use agreement. The population-wide cohort was identified from the Canadian Institute for Health Information Discharge Abstract Database, which contains diagnostic, intervention, and care information on all Ontario inpatient hospital stays. The cohort included all individuals who had an inpatient hospital stay from the 2002 through 2010 fiscal years in Ontario, Canada. Canadian Institutes for Health Information Discharge Abstract Database data were linked to the National Rehabilitation Reporting System, which captures admission, discharge, and treatment episode information on all IR stays province-wide.
      • Canadian Institute for Health Information
      National Rehabilitation Reporting System data quality documentation, 2007–2008.

       Case definition

      Patients with HIBI aged ≥20 years were identified in the Canadian Institutes for Health Information Discharge Abstract Database by the presence of an International Classification of Diseases, Tenth Revision (ICD-10) code for anoxic brain damage (G93.1) as the most responsible diagnosis. Patients with anoxic brain damage as a secondary diagnosis were additionally included given it was coincident with a most responsible diagnosis indicative of probable causative conditions, namely cardiac (I46.0) and respiratory arrest or asphyxia (R09.0, R09.2, T71, R09.0, and T75.1) and conditions likely to involve anoxia (G92, T58, and T70.2).

       Predisposing variables of interest

      Potential predictors of admission to IR were chosen based on previous research on determinants of acute care discharge destination and referral to rehabilitation among ABI populations
      • Jourdan C.
      • Bayen E.
      • Bosserelle V.
      • et al.
      Referral to rehabilitation after severe traumatic brain injury: results from the PariS-TBI Study.
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      and data availability of the Canadian Institutes for Health Information Discharge Abstract Database and National Rehabilitation Reporting System. Sociodemographic variables included age, sex, income quintile, and rurality, the latter 2 based on dissemination area, the smallest divisible geographic Canadian census unit. Prior extent of health care utilization and derived comorbidity burden over the 2 years prior to the acute care episode were captured using Johns Hopkins Aggregated Diagnosis Groups. Health care utilization was measured using Adjusted Clinical Groups Resource Utilization Band score,
      • Zielinski A.
      • Kronogard M.
      • Lenhoff H.
      • Halling A.
      Validation of ACG Case-mix for equitable resource allocation in Swedish primary health care.
      • Brilleman S.L.
      • Salisbury C.
      Comparing measures of multimorbidity to predict outcomes in primary care: a cross-sectional study.
      and a comorbidity index was derived from a weighted sum of Aggregated Diagnosis Groups accumulated over the 2 years prior to the admission for the HIBI acute care episode.
      • Austin P.C.
      • van Walraven C.
      • Wodchis W.P.
      • Newman A.
      • Anderson G.M.
      Using the Johns Hopkins Aggregated Diagnosis Groups (ADGs) to predict mortality in a general adult population cohort in Ontario, Canada.
      Acute care episode characteristics included comorbidity burden, length of stay (LOS) and hours spent in a special care unit, psychiatric comorbidity, and number of alternate level of care (ALC) days during the preceding acute care episode. Psychiatric comorbidities were defined as any mental health or behavioral disorder diagnosis (ICD-10 F chapter heading) in any secondary Canadian Institutes for Health Information Discharge Abstract Database diagnostic field.
      • Chen A.
      • Zagorski B.
      • Parsons D.
      • Vander Laan R.
      • Colantonio A.
      Acute care alternate-level-of-care days due to delayed discharge for traumatic and non-traumatic brain injuries.
      Number of ALC days were included as a marker of acute care delayed discharge, defined as days on which the attending physician deems the patient no longer requires acute care services while occupying an acute care bed.
      • Canadian Institute for Health Information
      Analysis in brief: alternate level of care in Canada.
      ALC days exclude those spent in special care units. Additionally, we considered fiscal year of acute care discharge to capture trends in IR admission with time and adjust for associated residual confounding.

       Main outcome

      The main outcome of interest was admission to IR. We additionally assessed admitting diagnosis (Diagnostic Health Conditions up to 2008 and by ICD-10-Canadian Enhancement from 2008 onward), rehabilitation program designation (Rehabilitation Client Group), and FIM.
      • Ottenbacher K.J.
      • Hsu Y.
      • Granger C.V.
      • Fiedler R.C.
      The reliability of the Functional Independence Measure: a quantitative review.
      • Canadian Institute for Health Information
      About National Rehabilitation Reporting System (NRS) QuickStats.

       Analyses

      Multivariable Poisson regression with robust variance estimation was used to estimate adjusted relative incidence proportions (relative risks) of entry to IR across potential acute care–specific and demographic predictors of interest.
      • Zou G.
      A modified Poisson regression approach to prospective studies with binary data.
      • Spiegelman D.
      • Hertzmark E.
      Easy SAS calculations for risk or prevalence ratios and differences.
      Poisson regression models were selected over log binomial models because of questionable convergence encountered during model building.
      Model building for multivariable regression began by examining bivariable associations between each potential predictor and entry to IR. Variables were not retained in the final multivariable model when maximum likelihood-based diagnostics of model fit (eg, likelihood ratio test) indicated that they did not contribute to the model. Where appropriate, categorical variables were collapsed to provide the most informative description of effects across levels. For all analyses, HIBI acute care survivors remained eligible for entry to IR for up to 1 year posthospital discharge. In the rare case of multiple subsequent IR episodes per individual, only the most proximal was included. SAS version 9.3a was used for all analyses.

      Results

      There were 599 patients with HIBI who survived their index acute care episode over the study period. Of these, 75% had anoxic brain damage as the most responsible diagnosis for their preceding acute care episode. Of the 599 HIBI acute care survivors, 180 patients had at least 1 subsequent admission to IR; of these, 93.9% occurred within a year of acute care discharge, and 64.4% were admitted within 1 day of acute care discharge (ie, direct transfer).
      Of those who were admitted to IR between 1 day and 1 year postacute discharge (n=169), the median time since admission to acute care was 44 days (interquartile range, 27–84). Over half of this group (56.2%) had an anoxic brain damage admitting diagnosis in IR. Of the 169, 69.2% were designated for ABI-specific rehabilitation (admitting Rehabilitation Client Groups: 02.x). Mean total and cognitive FIM summary scores were 69.8±30.9 and 18.3±8.53, respectively, at IR admission.
      Six patients were omitted because of missing data; the remaining 593 comprised the study sample. The distribution of potential demographic and acute care predictors across the overall sample, and those admitted to IR within 1 year, are summarized in table 1. Independent sociodemographic and preacute care predictors of IR admission were age, sex, and comorbidity burden (table 2). Relative to patients aged ≥80 years, those <65 years were statistically significantly more likely to be admitted to IR (P trend, <.001). Women were substantially less likely to be admitted to IR than men (relative risk, .62; 95% confidence interval, .46–.84). An Aggregated Diagnosis Group comorbidity score >30 coincided with a 33% lower likelihood of IR admission (relative risk, .67; 95% confidence interval, .46–.97) relative to those with an Aggregated Diagnosis Group score <10. Socioeconomic status and rurality were not significantly associated with IR admission.
      Table 1Distribution of potential determinants of admission to IR among all acute care HIBI survivors and only those admitted to IR
      Potential DeterminantsHIBI Survivors (N=593)IR Within 1y (n=169)
      Sociodemographic and comorbidity
       Age (y)
      20–3454 (9.1)24 (14.2)
      35–49107 (18.0)44 (26.0)
      50–64189 (31.9)54 (32.0)
      65–79188 (31.7)41 (24.3)
      ≥8055 (9.3)6 (3.6)
       Sex
      Female204 (34.4)40 (23.7)
      Male389 (65.6)129 (76.3)
       Income quintiles
      Quintile 1 (lowest)165 (27.8)40 (23.7)
      Quintile 2126 (21.2)43 (25.4)
      Quintile 3105 (17.7)31 (18.3)
      Quintile 4123 (20.7)35 (20.7)
      Quintile 5 (highest)74 (12.5)20 (11.8)
       Rurality
      Rural58 (9.8)19 (11.2)
      Urban535 (90.2)150 (88.8)
       ACG RUB
      None or healthy34 (5.7)15 (8.9)
      Low35 (5.9)12 (7.1)
      Moderate198 (33.4)56 (33.1)
      High152 (25.6)45 (26.6)
      Very high174 (29.3)41 (24.3)
       ADG score
      <10187 (31.5)66 (39.0)
      10–30265 (44.7)75 (44.4)
      >30141 (23.8)28 (16.6)
      Prior acute care episode
       LOS (d)
      <10119 (20.1)12 (7.1)
      10–30203 (34.2)55 (32.5)
      31–90210 (35.4)84 (49.7)
      >9061 (10.3)18 (10.6)
       Time in SCU (h)
      None86 (14.5)23 (13.6)
      1–999428 (72.2)124 (73.4)
      >100079 (13.3)22 (13.0)
       Delayed discharge (ALC days
      ALC days are an indicator of delayed discharge from acute care, defined as the number of days on which the attending physician deems that appropriate care can no longer be provided.
      )
      0354 (59.7)81 (47.9)
      1–14100 (16.9)43 (25.4)
      >14149 (23.4)45 (25.6)
       Psychological/behavioral
      Absent452 (76.2)126 (74.6)
      Present141 (23.8)43 (25.4)
       Fiscal years
      2002–2003137 (23.1)29 (17.2)
      2004–2005123 (20.7)37 (21.9)
      2006–2007131 (22.1)34 (20.1)
      2008–2009138 (23.3)48 (28.4)
      NOTE. Values are n (%).
      Abbreviations: ACG, Adjusted Clinical Group; ADG, Aggregated Diagnosis Group; RUB, Resource Utilization Band; SCU, special care unit.
      ALC days are an indicator of delayed discharge from acute care, defined as the number of days on which the attending physician deems that appropriate care can no longer be provided.
      Table 2Independent effects of predictors of admission to IR represented as RRs and 95% CIs among 593 acute care survivors of HIBI
      Potential DeterminantsRR95% CI
      Age (y)
       20–343.811.70–8.53
       35–493.571.64–7.79
       50–642.471.14–5.37
       65–791.890.86–4.15
       ≥80 (ref)1.00NA
      Sex
       Female0.620.46–0.84
       Male1.00
      ADG score
       <10 (ref)1.00NA
       10–300.910.70–1.18
       >300.670.46–0.97
      LOS in acute care (d)
       <10 (ref)1.00NA
       10–302.461.40–4.31
       31–904.012.24–7.18
       >903.441.65–7.18
      Time in SCU (h)
       01.00NA
       1–9990.790.54–1.15
       ≥10000.500.30–0.83
      Delayed discharge (ALC days
      ALC days are an indicator of delayed discharge from acute care, defined as the number of days on which the attending physician deems that appropriate care can no longer be provided.
      )
       01.00NA
       1–141.501.09–2.06
       >140.890.63–1.28
      Fiscal years
       2002–20030.630.39–1.01
       2004–20050.870.56–1.37
       2006–20070.720.46–1.13
       2008–20091.060.70–1.60
       2010–2012 (ref)1.00NA
      Abbreviations: ADG, Aggregated Diagnosis Group; CI, confidence interval; NA, not applicable; ref, reference; RR, relative incidence proportion; SCU, special care unit.
      ALC days are an indicator of delayed discharge from acute care, defined as the number of days on which the attending physician deems that appropriate care can no longer be provided.
      Independent predictors of admission to IR associated with the preceding HIBI acute care episode included LOS, accumulated time in special care, and having a moderately delayed discharge indicated by number of ALC days (see table 2). Longer LOS was associated with an increased likelihood of IR admission (P trend, <.001); however, the point estimate for the highest category (LOS >90d) decreased over the preceding contiguous category. Patients with HIBI who spent ≥1000 hours in special care were half as likely to end up in IR relative to those with no special care hours (relative risk, .50; 95% confidence interval, .30–.83). Finally, relative to those without delayed discharge from acute care, spending 1 day to 2 weeks in an ALC bed was associated with a 50% increased probability of admission to IR (relative risk, 1.50; 95% confidence interval, 1.09–2.06).

      Discussion

      Using a population-wide Ontario sample, our findings demonstrated that approximately 30% of HIBI survivors were admitted to IR within 1 year of discharge from acute care and that over two thirds of those were transferred directly. Of patients with HIBI admitted to IR within 1 year, most received rehabilitation specific to ABI. Significant sociodemographic and preacute care predictors of admission to IR included younger age, male sex, and lower comorbidity burden. Acute care specific predictors comprised longer LOS, shorter time in special care, and moderately delayed discharge. That income quintile and rurality were not associated with admission to IR may suggest that access to such services among HIBI survivors in Ontario, Canada, is not affected substantially by socioeconomic status.
      This article found that 28% of HIBI survivors were admitted to IR within 1 year of acute care discharge. This contrasts with findings on other brain injury populations, where 9.8% of TBI and 8.8% of NTBI survivors were found to be discharged to IR.
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      Although this may be caused by the direct linkage of inpatient hospital and IR records and a longer follow-up period (1y vs direct discharge), the proportion of HIBI survivors going to IR in this study is in line with those of patients with more severe TBI seen in trauma centers (27.6%).
      • de Guise E.
      • Feyz M.
      • LeBlanc J.
      • Richard S.L.
      • Lamoureux J.
      Overview of traumatic brain injury patients at a tertiary trauma centre.
      Similarly, of 211 cardiac arrest survivors in the United States, 16% were discharged to an acute rehabilitation facility.
      • Rittenberger J.C.
      • Raina K.
      • Holm M.B.
      • Kim Y.J.
      • Callaway C.W.
      Association between cerebral performance category, modified Rankin scale, and discharge disposition after cardiac arrest.
      Nevertheless, it is clear that a significant proportion of patients with HIBI use IR; however, there is a paucity of information on this population. Given the high costs of IR,
      • Chen A.
      • Bushmeneva K.
      • Zagorski B.
      • Colantonio A.
      • Parsons D.
      • Wodchis W.
      Direct cost associated with acquired brain injury in Ontario.
      and the existing sparse evidence that patients with HIBI experience less benefit from IR relative to TBI,
      • Cullen N.K.
      • Park Y.
      • Bayley M.T.
      Functional recovery following traumatic vs. non-traumatic brain injury: a case-controlled study.
      • Cullen N.
      • Crescini C.
      • Bailey M.
      Rehabilitation outcomes after anoxic brain injury: a case controlled comparison with traumatic brain injury.
      • Cullen N.K.
      • Weisz K.
      Cognitive correlates with functional outcomes after anoxic brain injury: a case-controlled comparison with traumatic brain injury.
      additional research on patients with HIBI in the IR setting is needed to inform service planning and resource allocation for this group.
      This study found that, among those admitted to IR, mean total and cognitive FIM summary scores at admission were 69.8±30.9 and 18.3±8.53, respectively. A recent population-wide study on patients with ABI receiving IR services in Canada found that the mean FIM scores for patients with NTBI and TBI at admission were 76.8±26.5 and 79.4±28.9, respectively.
      • Colantonio A.
      • Gerber G.
      • Bayley M.
      • Deber R.
      • Yin J.
      • Kim H.
      Differential profiles for patients with traumatic and non-traumatic brain injury.
      Similarly, mean cognitive FIM scores at admission to IR were 21.8 for NTBI and 21 for TBI. In contrast, our data on patients with HIBI suggest that they represent a more severe patient population. This is important given that these patient groups are often treated in the same rehabilitation setting.
      • Cullen N.
      • Crescini C.
      • Bailey M.
      Rehabilitation outcomes after anoxic brain injury: a case controlled comparison with traumatic brain injury.
      Further, it is reasonable to assume that lower FIM scores correspond to a greater burden of care.
      This study found that, of those admitted to IR, just over half had an HIBI-associated admitting diagnosis, and 69.2% received rehabilitation specific to ABI. This finding raises concerns regarding the quality of HIBI patient care in IR. For example, approximately 15% of this cohort had a Rehabilitation Client Group indicating rehabilitation for cardiac arrest; however, cardiac rehabilitation does not necessarily address cognitive impairments, typical sequelae of HIBI.
      • Wilson B.A.
      Cognitive functioning of adult survivors of cerebral hypoxia.
      Although discrepancies between diagnosis and designation of rehabilitative programming may partially explain reduced effectiveness of IR after HIBI,
      • Cullen N.K.
      • Park Y.
      • Bayley M.T.
      Functional recovery following traumatic vs. non-traumatic brain injury: a case-controlled study.
      • Cullen N.
      • Crescini C.
      • Bailey M.
      Rehabilitation outcomes after anoxic brain injury: a case controlled comparison with traumatic brain injury.
      • Cullen N.K.
      • Weisz K.
      Cognitive correlates with functional outcomes after anoxic brain injury: a case-controlled comparison with traumatic brain injury.
      • Smania N.
      • Avesani R.
      • Roncari L.
      • et al.
      Factors predicting functional and cognitive recovery following severe traumatic, anoxic, and cerebrovascular brain damage.
      the extent to which the needs of persons with HIBI are met by IR programing not designed specifically for ABI warrants further investigation. Nevertheless, it is important that needed cognitive rehabilitation is provided to HIBI survivors that enter non–ABI-specific rehabilitation units.
      A key finding was that women were almost 2-fold less likely to be admitted to IR. Women have been found to be significantly less likely to be referred to, and enrolled in, cardiac rehabilitation,
      • Samayoa L.
      • Grace S.L.
      • Gravely S.
      • Scott L.B.
      • Marzolini S.
      • Colella T.J.
      Sex differences in cardiac rehabilitation enrollment: a meta-analysis.
      • Stewart Williams J.A.
      Using non-linear decomposition to explain the discriminatory effects of male-female differentials in access to care: a cardiac rehabilitation case study.
      to be admitted to IR after stroke,
      • Sandel M.E.
      • Wang H.
      • Terdiman J.
      • et al.
      Disparities in stroke rehabilitation: results of a study in an integrated health system in northern California.
      and to be admitted to neurorehabilitation after subarachnoid hemorrhage.
      • Macleod M.R.
      • Smith S.J.
      Gender and deprivation and rates of referral and thereby admission to a national neurorehabilitation service.
      However, no sex differences in rate of admission to rehabilitation after TBI have been documented.
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      • Macleod M.R.
      • Smith S.J.
      Gender and deprivation and rates of referral and thereby admission to a national neurorehabilitation service.
      Research has not demonstrated a sex difference in benefit from rehabilitation after brain injury,
      • Graham J.E.
      • Radice-Neumann D.M.
      • Reistetter T.A.
      • Hammond F.M.
      • Dijkers M.
      • Granger C.V.
      Influence of sex and age on inpatient rehabilitation outcomes among older adults with traumatic brain injury.
      • Ratcliff J.J.
      • Greenspan A.I.
      • Goldstein F.C.
      • et al.
      Gender and traumatic brain injury: do the sexes fare differently?.
      which implies that the choice of referral and access to IR should not be influenced by sex. In explanation, it is possible that, even in a country with universal access to health care, rehabilitation overemphasizes return to work as a treatment goal and may be disproportionately allocated to the predominantly male population in paid work.
      • Macleod M.R.
      • Smith S.J.
      Gender and deprivation and rates of referral and thereby admission to a national neurorehabilitation service.
      • Altman B.M.
      • Smith R.T.
      Rehabilitation service utilization models: changes in the opportunity structure for disabled women.
      • Haag H.L.
      • Caringal M.
      • Sokoloff S.
      • Kontos P.
      • Yoshida K.
      • Colantonio A.
      Being a woman with acquired brain injury: challenges and implications for practice.
      Unfortunately, we were unable to ascertain definitive employment status prior to injury for all participants. Additionally, barriers that limit enrollment of women in rehabilitation have been suggested to include lack of confidence in the ability to benefit, a perceived imperative to return to a care-providing role in the home, or that women are more resourceful and generally require less rehabilitation services.
      • Macleod M.R.
      • Smith S.J.
      Gender and deprivation and rates of referral and thereby admission to a national neurorehabilitation service.
      Health care providers have reported that women tend to benefit less from advocacy from family members when they are ill.
      • Haag H.L.
      • Caringal M.
      • Sokoloff S.
      • Kontos P.
      • Yoshida K.
      • Colantonio A.
      Being a woman with acquired brain injury: challenges and implications for practice.
      Measures of social support beyond those available from inpatient hospital administrative data are advocated for investigation in future studies. A further proposition is that the cause and severity of HIBI differs by sex. In support of this, a recent study on 1469 patients with ABI in Italy found men more likely to suffer HIBI with anoxic pathology (eg, cardiac arrest) rather than vascular (eg, subarachnoid and intraparenchymal hemorrhage, ischemia) or other pathology.
      • Avesvani R.
      • Roncari L.
      • Khansefid M.
      • et al.
      The Italian National Registry of severe acquired brain injury: epidemiological, clinical and functional data of 1469 patients.
      Patients with anoxic pathology presented the highest functional impairment at rehabilitation admission.
      • Avesvani R.
      • Roncari L.
      • Khansefid M.
      • et al.
      The Italian National Registry of severe acquired brain injury: epidemiological, clinical and functional data of 1469 patients.
      If women are more likely to be denied admission to IR, despite having equal need, it is reasonable to speculate that they have worse long-term prognoses after surviving a HIBI. Unfortunately we did not have access to direct measures of need (eg, injury severity) in the administrative health data used for this study. Further investigation of disparities in access to care among HIBI survivors is recommended to characterize any sex or sex inequities that may exist, using both quantitative and qualitative methods.
      Our results, in line with previous research on TBI, demonstrate that the probability of admission to IR decreases with age.
      • Jourdan C.
      • Bayen E.
      • Bosserelle V.
      • et al.
      Referral to rehabilitation after severe traumatic brain injury: results from the PariS-TBI Study.
      • Foster M.
      • Tilse C.
      • Fleming J.
      Referral to rehabilitation following traumatic brain injury: practitioners and the process of decision-making.
      We also found an inverse relation between comorbidity burden and IR admission, as reported previously.
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      It is less clear how severity of HIBI, as indicated by special care unit hours and LOS in acute care, affects likelihood of IR admission. Those that had accumulated >1000 special care unit hours were half as likely to be admitted to IR relative to those that did not accumulate any. This is in opposition with prior research from France indicating LOS in intensive care to be positively associated with referral to rehabilitation in patients with TBI.
      • Jourdan C.
      • Bayen E.
      • Bosserelle V.
      • et al.
      Referral to rehabilitation after severe traumatic brain injury: results from the PariS-TBI Study.
      We did, however, observe a strong positive association between LOS in the prior HIBI acute care episode and admission to IR, as found previously with TBI,
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      suggesting it is a better indicator of injury severity than special care unit time. Overall, these findings suggest that cost-reduction measures may play a role in determining access to IR among Canadian patients with HIBI because lower patient age and lower comorbidity burden are positively associated with admission to IR, and these factors have been associated with lower rehabilitation costs and shorter rehabilitation LOS.
      • Graham J.E.
      • Radice-Neumann D.M.
      • Reistetter T.A.
      • Hammond F.M.
      • Dijkers M.
      • Granger C.V.
      Influence of sex and age on inpatient rehabilitation outcomes among older adults with traumatic brain injury.
      • Bejor M.
      • Ramella F.C.
      • Toffola E.D.
      • Comelli M.
      • Chiappedi M.
      Inpatient rehabilitation outcome: a matter of diagnosis?.
      Alternatively, it may be that patients who spend >1000 hours in the special care unit sustain more severe injuries, have worse prognoses, and are deemed unlikely to benefit from rehabilitation and are instead discharged to continuing/long-term/palliative care.
      • Chen A.Y.
      • Zagorski B.
      • Parsons D.
      • et al.
      Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
      Finally, we found an association between moderately delayed acute care discharge and admission to IR. Canadian Institute for Health Information data have demonstrated that 43% of patients receiving ALC in 2007 to 2008 were awaiting discharge to a long-term care facility and 13% were awaiting discharge to a rehabilitation facility.
      • Zielinski A.
      • Kronogard M.
      • Lenhoff H.
      • Halling A.
      Validation of ACG Case-mix for equitable resource allocation in Swedish primary health care.
      Unpublished results from related prior work (Stock D, Cowie C, Chan V, et al, unpublished data, 2015) have corroborated that delayed discharge among patients with HIBI is strongly associated with waiting for palliative or long-term care in Ontario and much less so for IR. The increased likelihood of admission to IR for those who accumulated 1 to 14 ALC days in the preceding HIBI acute care episode may be attributed to patients with HIBI deemed eligible to benefit from IR but unfit to be discharged home, together with wait times for IR being relatively modest. Prior research on the TBI population has found that for every 4 to 5 days of delayed admission to rehabilitation, total FIM and motor FIM scores decrease on average by 1 point.
      • Kunik C.L.
      • Flowers L.
      • Kazanjian T.
      Time to rehabilitation admission and associated outcomes for patients with traumatic brain injury.
      Patients with HIBI waiting for IR may therefore still be needlessly using valuable acute care beds as they await admission, which could adversely affect rehabilitation outcomes.
      This study has considerable strengths. It is the first, to our knowledge, to examine predictors of IR admission among patients with HIBI. Given that Ontario represents approximately 40% of the Canadian population, these findings are highly generalizable.
      • Statistics Canada
      Population by year, by province and territory.
      Data quality assessments support coding in the Canadian Institutes for Health Information Discharge Abstract Database and Canadian Institutes for Health Information National Rehabilitation Reporting System to be valid.
      • Juurlink D.
      • Preyra C.
      • Croxford R.
      • et al.
      Canadian Institute for Health Information Discharge Abstract Database: a validation study.

      Canadian Institute for Health Information. Rehabilitation. Available at: https://www.cihi.ca/en/types-of-care/hospital-care/rehabilitation. Accessed June 8, 2015.

      Additionally, the large population-wide sample from these databases allowed multiple potential predictors to be evaluated. Finally, our study has the strength of linking records across the acute care–IR continuum.

      Study limitations

      These findings warrant consideration of certain limitations. First, we did not have a direct measure of HIBI severity in our data, and there is currently no such criterion standard measure. Further, research has found that social factors affect referral to rehabilitation (eg, living situation, alcohol abuse
      • Jourdan C.
      • Bayen E.
      • Bosserelle V.
      • et al.
      Referral to rehabilitation after severe traumatic brain injury: results from the PariS-TBI Study.
      ); however, we were not able to examine such factors given their absence in our inpatient hospital administrative health data. Third, there is no empirically validated standard clinical case definition or formal diagnostic criteria for HIBI. The ICD-10 code for anoxic brain damage may connote too strongly its severity, potentially failing to capture cases that do not involve anoxia, but instead hypoxia or hypoxia-ischemia.
      • Arciniegas D.B.
      Hypoxic-ischemic brain injury: addressing the disconnect between pathophysiology and public policy.
      We attempted to increase case sensitivity by additionally including anoxic brain damage (ICD-10 code G93.1) as a secondary acute care diagnosis coincident with a primary diagnosis indicative of a causal HIBI mechanism. However, it is possible that a substantial proportion of milder, clinically relevant HIBI cases were missed.

      Conclusions

      Our findings suggest a significant proportion of patients are admitted to IR after survival of HIBI with a more severe functional impairment level relative to other ABIs. Future research on the HIBI IR population should evaluate whether significant gains are made and how they compare to other ABI populations. To date, this has only been investigated in small-sample case-control studies. Although these studies have suggested that patients with HIBI have a slower, less marked recovery,
      • Cullen N.K.
      • Park Y.
      • Bayley M.T.
      Functional recovery following traumatic vs. non-traumatic brain injury: a case-controlled study.
      • Cullen N.
      • Crescini C.
      • Bailey M.
      Rehabilitation outcomes after anoxic brain injury: a case controlled comparison with traumatic brain injury.
      • Cullen N.K.
      • Weisz K.
      Cognitive correlates with functional outcomes after anoxic brain injury: a case-controlled comparison with traumatic brain injury.
      • Smania N.
      • Avesani R.
      • Roncari L.
      • et al.
      Factors predicting functional and cognitive recovery following severe traumatic, anoxic, and cerebrovascular brain damage.
      this needs to be examined in larger population-based cohorts.
      Independent predictors of admission to IR indicate that admission favors those who will likely require less expensive services and have a shorter LOS. Although some HIBI survivors may be too severely impaired to benefit from IR, policymakers should nevertheless focus on expanding current IR programming to effectively serve a wider range of disability in the interest of maximizing equity of care. Finally, future research should examine causes behind the significantly reduced likelihood of admission to IR experienced by female survivors of HIBI.

      Supplier

      • a.
        SAS version 9.3; SAS Institute.

      Supplementary data

      References

        • Toronto ABI Network
        Definition of ABI.
        2015 (Available at:) (Accessed June 9, 2015)
        • Basso A.
        • Previgliano I.
        • Servadei F.
        Neurological disorders: a public health approach. Traumatic brain injuries.
        in: Neurological disorders: public health challenges. World Health Organization, 2006: 164-176
        • Ropper A.H.
        • Samuels M.A.
        The acquired metabolic disorders of the nervous system.
        in: Ropper A.H. Samuels M.A. Adams and Victor’s principles of neurology. McGraw Hill Medical, New York2009: 1081-1107
        • Arciniegas D.B.
        Hypoxic-ischemic brain injury: addressing the disconnect between pathophysiology and public policy.
        Neurorehabilitation. 2010; 26: 1-4
      1. Colantonio A, Chan V, Zagorski B, Parsons D, Vander Laan R. Ontario Acquired Brain Injury (ABI) Dataset Project Phase III: highlights: number of episodes of care and causes of brain injury. Available at: http://www.abiresearch.utoronto.ca/ABI%20Dataset%20+%20LHIN%20USE%20THIS/LHIN%20Factsheets%20Demographics%20-%20June%206.%202013%20-%20REVISED%20WITH%20NEW%20LINK.pdf. Accessed June 17, 2015.

        • Wilson B.A.
        Cognitive functioning of adult survivors of cerebral hypoxia.
        Brain Inj. 1996; 10: 863-874
        • Lu-Emerson C.
        • Khot S.
        Neurological sequelae of hypoxic-ischemic brain injury.
        Neurorehabilitation. 2010; 26: 35-45
        • Khot S.
        • Tirschwell D.L.
        Long-term neurological complications after hypoxic-ischemic encephalopathy.
        Semin Neurol. 2006; 26: 422-431
        • Wilson M.
        • Staniforth A.
        • Till R.
        • das Nair R.
        • Vesey P.
        The psychosocial outcomes of anoxic brain injury following cardiac arrest.
        Resuscitation. 2014; 85: 795-800
        • Middelkamp W.
        • Moulaert V.
        • Verbunt J.A.
        • van Heugten C.M.
        • Bakx W.G.
        • Wade D.T.
        Life after survival: long-term daily life functioning and quality of life of patients with hypoxic injury as a result of cardiac arrest.
        Clin Rehabil. 2007; 21: 425-431
        • Cullen N.K.
        • Park Y.
        • Bayley M.T.
        Functional recovery following traumatic vs. non-traumatic brain injury: a case-controlled study.
        Brain Inj. 2008; 22: 1013-1020
        • Cullen N.
        • Crescini C.
        • Bailey M.
        Rehabilitation outcomes after anoxic brain injury: a case controlled comparison with traumatic brain injury.
        PMR. 2009; 1: 1069-1076
        • Cullen N.K.
        • Weisz K.
        Cognitive correlates with functional outcomes after anoxic brain injury: a case-controlled comparison with traumatic brain injury.
        Brain Inj. 2011; 25: 35-43
        • Smania N.
        • Avesani R.
        • Roncari L.
        • et al.
        Factors predicting functional and cognitive recovery following severe traumatic, anoxic, and cerebrovascular brain damage.
        J Head Trauma Rehabil. 2013; 28: 131-140
        • Fitzgerald A.
        • Aditya H.
        • Prior A.
        • McNeill E.
        • Pentland B.
        Anoxic brain injury: clinical patterns and functional outcomes: a study of 93 cases.
        Brain Inj. 2010; 24: 1311-1323
        • Fertl E.
        • Vass K.
        • Sterz F.
        • Gabriel H.
        • Auf E.
        Neurological rehabilitation of severely disabled cardiac arrest survivors. Part 1. Course of post-acute inpatient treatment.
        Resuscitation. 2000; 47: 231-239
        • Chan V.
        • Zagorski B.
        • Parsons D.
        • Colantonio A.
        Older adults with acquired brain injury: a population based study.
        BMC Geriatr. 2013; 13: 97
        • Kraus J.F.
        • Chu L.D.
        Epidemiology.
        in: Silver J.M. McAllister T.M. Yudofsky S.C. Textbook of traumatic brain injury. Arlington: American Psychiatric Publishing, 2005: 3-26
        • Jourdan C.
        • Bayen E.
        • Bosserelle V.
        • et al.
        Referral to rehabilitation after severe traumatic brain injury: results from the PariS-TBI Study.
        Neurorehabil Neural Repair. 2013; 27: 35-44
        • Foster M.
        • Tilse C.
        • Fleming J.
        Referral to rehabilitation following traumatic brain injury: practitioners and the process of decision-making.
        Soc Sci Med. 2004; 59: 1867-1878
        • Chen A.Y.
        • Zagorski B.
        • Parsons D.
        • et al.
        Factors associated with discharge destination from acute care after acquired brain injury in Ontario, Canada.
        BMC Neurol. 2012; 12: 16
        • Canadian Institute for Health Information
        National Rehabilitation Reporting System data quality documentation, 2007–2008.
        CIHI, Ottawa2009
        • Zielinski A.
        • Kronogard M.
        • Lenhoff H.
        • Halling A.
        Validation of ACG Case-mix for equitable resource allocation in Swedish primary health care.
        BMC Public Health. 2009; 9: 347
        • Brilleman S.L.
        • Salisbury C.
        Comparing measures of multimorbidity to predict outcomes in primary care: a cross-sectional study.
        Fam Pract. 2013; 30: 172-178
        • Austin P.C.
        • van Walraven C.
        • Wodchis W.P.
        • Newman A.
        • Anderson G.M.
        Using the Johns Hopkins Aggregated Diagnosis Groups (ADGs) to predict mortality in a general adult population cohort in Ontario, Canada.
        Med Care. 2011; 49: 932-939
        • Chen A.
        • Zagorski B.
        • Parsons D.
        • Vander Laan R.
        • Colantonio A.
        Acute care alternate-level-of-care days due to delayed discharge for traumatic and non-traumatic brain injuries.
        Healthc Policy. 2012; 7: 41-55
        • Canadian Institute for Health Information
        Analysis in brief: alternate level of care in Canada.
        2009 (Available at:) (Accessed May 30, 2015)
        • Ottenbacher K.J.
        • Hsu Y.
        • Granger C.V.
        • Fiedler R.C.
        The reliability of the Functional Independence Measure: a quantitative review.
        Arch Phys Med Rehabil. 1996; 77: 1226-1232
        • Canadian Institute for Health Information
        About National Rehabilitation Reporting System (NRS) QuickStats.
        2014 (Available at:) (Accessed June 5, 2015)
        • Zou G.
        A modified Poisson regression approach to prospective studies with binary data.
        Am J Epidemiol. 2004; 159: 702-706
        • Spiegelman D.
        • Hertzmark E.
        Easy SAS calculations for risk or prevalence ratios and differences.
        Am J Epidemiol. 2005; 162: 199-200
        • de Guise E.
        • Feyz M.
        • LeBlanc J.
        • Richard S.L.
        • Lamoureux J.
        Overview of traumatic brain injury patients at a tertiary trauma centre.
        Can J Neurol Sci. 2005; 32: 186-193
        • Rittenberger J.C.
        • Raina K.
        • Holm M.B.
        • Kim Y.J.
        • Callaway C.W.
        Association between cerebral performance category, modified Rankin scale, and discharge disposition after cardiac arrest.
        Resuscitation. 2011; 82: 1036-1040
        • Chen A.
        • Bushmeneva K.
        • Zagorski B.
        • Colantonio A.
        • Parsons D.
        • Wodchis W.
        Direct cost associated with acquired brain injury in Ontario.
        BMC Neurol. 2012; 12: 76
        • Colantonio A.
        • Gerber G.
        • Bayley M.
        • Deber R.
        • Yin J.
        • Kim H.
        Differential profiles for patients with traumatic and non-traumatic brain injury.
        J Rehabil Med. 2011; 43: 311-315
        • Samayoa L.
        • Grace S.L.
        • Gravely S.
        • Scott L.B.
        • Marzolini S.
        • Colella T.J.
        Sex differences in cardiac rehabilitation enrollment: a meta-analysis.
        Can J Cardiol. 2014; 30: 793-800
        • Stewart Williams J.A.
        Using non-linear decomposition to explain the discriminatory effects of male-female differentials in access to care: a cardiac rehabilitation case study.
        Soc Sci Med. 2009; 69: 1072-1079
        • Sandel M.E.
        • Wang H.
        • Terdiman J.
        • et al.
        Disparities in stroke rehabilitation: results of a study in an integrated health system in northern California.
        PMR. 2009; 1: 29-40
        • Macleod M.R.
        • Smith S.J.
        Gender and deprivation and rates of referral and thereby admission to a national neurorehabilitation service.
        Clin Rehabil. 2005; 19: 109-115
        • Graham J.E.
        • Radice-Neumann D.M.
        • Reistetter T.A.
        • Hammond F.M.
        • Dijkers M.
        • Granger C.V.
        Influence of sex and age on inpatient rehabilitation outcomes among older adults with traumatic brain injury.
        Arch Phys Med Rehabil. 2010; 91: 43-50
        • Ratcliff J.J.
        • Greenspan A.I.
        • Goldstein F.C.
        • et al.
        Gender and traumatic brain injury: do the sexes fare differently?.
        Brain Inj. 2007; 21: 1023-1030
        • Altman B.M.
        • Smith R.T.
        Rehabilitation service utilization models: changes in the opportunity structure for disabled women.
        Int Disabil Stud. 1990; 12: 149-156
        • Haag H.L.
        • Caringal M.
        • Sokoloff S.
        • Kontos P.
        • Yoshida K.
        • Colantonio A.
        Being a woman with acquired brain injury: challenges and implications for practice.
        Arch Phys Med Rehabil. 2016; 97: S64-S70
        • Avesvani R.
        • Roncari L.
        • Khansefid M.
        • et al.
        The Italian National Registry of severe acquired brain injury: epidemiological, clinical and functional data of 1469 patients.
        Eur J Phys Rehabil Med. 2013; 49: 611-618
        • Bejor M.
        • Ramella F.C.
        • Toffola E.D.
        • Comelli M.
        • Chiappedi M.
        Inpatient rehabilitation outcome: a matter of diagnosis?.
        Neuropsychiatr Dis Treat. 2013; 9: 253-257
        • Kunik C.L.
        • Flowers L.
        • Kazanjian T.
        Time to rehabilitation admission and associated outcomes for patients with traumatic brain injury.
        Arch Phys Med Rehabil. 2006; 87: 1590-1596
        • Statistics Canada
        Population by year, by province and territory.
        2014 (Available at:) (Accessed June 8, 2015)
        • Juurlink D.
        • Preyra C.
        • Croxford R.
        • et al.
        Canadian Institute for Health Information Discharge Abstract Database: a validation study.
        Institute for Clinical Evaluative Sciences, Toronto2006
      2. Canadian Institute for Health Information. Rehabilitation. Available at: https://www.cihi.ca/en/types-of-care/hospital-care/rehabilitation. Accessed June 8, 2015.