WHILE THERE IS CLEAR evidence that the average

Transcription

1 ORIGINAL ARTICLE Inpatient and Postdischarge Rehabilitation Services Provided in the First Year After Spinal Cord Injury: Findings From the SCIRehab Study Gale G. Whiteneck, PhD, Julie Gassaway, MS, RN, Marcel P. Dijkers, PhD, Daniel P. Lammertse, MD, Flora Hammond, MD, Allen W. Heinemann, PhD, Deborah Backus, PT, PhD, Susan Charlifue, PhD, Pamela H. Ballard, MD, Jeanne M. Zanca, MPT, PhD 361 ABSTRACT. Whiteneck GG, Gassaway J, Dijkers MP, Lammertse DP, Hammond F, Heinemann AW, Backus D, Charlifue S, Ballard PH, Zanca JM. Inpatient and postdischarge rehabilitation services provided in the first year after spinal cord injury: findings from the SCIRehab study. Arch Phys Med Rehabil 2011;92: Objective: To examine the amount and type of therapy services received in inpatient and postdischarge settings during the first year after spinal cord injury (SCI). Design: Prospective observational longitudinal cohort design. Data were obtained from systematic recording of interventions by clinicians and from patient interview. Setting: Inpatient and postdischarge rehabilitation programs. Participants: Patients (N 493) with traumatic SCI admitted to 6 rehabilitation centers participating in the SCIRehab study. Interventions: Not applicable. Main Outcome Measures: Hours of therapy by physical therapy (PT), occupational therapy (OT), speech therapy, recreation therapy, psychology, social work/case management, and nursing education during initial inpatient rehabilitation and postdischarge up to the first anniversary of injury. Inpatient data were collected prospectively by the treating clinicians; postdischarge service data were collected by patient self-report during follow-up interviews. Results: Of the total hours spent on these rehabilitation interventions during the first year after injury, 44% occurred after discharge from inpatient rehabilitation. Participants received 56% of their PT hours after discharge and 52% of their OT hours, but only a minority received any postdischarge From Craig Hospital, Englewood, CO (Whiteneck, Lammertse, Charlifue); Institute for Outcomes Research, Salt Lake City, UT (Gassaway); Mount Sinai School of Medicine, New York, NY (Dijkers, Zanca); Carolinas Rehabilitation, Charlotte, NC, and Indiana University, Indianapolis, IN (Hammond); Department of Physical Medicine and Rehabilitation, Feinberg School of Medicine, Northwestern University and Rehabilitation Institute of Chicago, Chicago, IL (Heinemann); Shepherd Center, Atlanta, GA (Backus); and National Rehabilitation Hospital, Washington, DC (Ballard). Supported by grants from the National Institute on Disability and Rehabilitation Research, Office of Rehabilitative Services, U.S. Department of Education, to Craig Hospital (grant nos. H133A and H133N060005), Mount Sinai School of Medicine (grant no. H133N060027), the National Rehabilitation Hospital (grant no. H133N060028), Rehabilitation Institute of Chicago (grant no. H133N060014), Shepherd Center (grant no. H133N060009), and Carolinas Rehabilitation. The opinions contained in this publication are those of the grantees and do not necessarily reflect those of the U.S. Department of Education. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Correspondence to Gale G. Whiteneck, PhD, Director of Research, 3425 S Clarkson St, Englewood, CO 80113, gale@craig-hospital.org /11/ $36.00/0 doi: /j.apmr services from other rehabilitation disciplines. While wide variation was found in the total hours of inpatient treatment across all disciplines, the variation in the total hours of postdischarge services was greater, with the interquartile range of postdischarge services being twice that of the inpatient services. Conclusions: SCI rehabilitation is often given in a care continuum, with inpatient rehabilitation being only the beginning. Reductions in inpatient SCI rehabilitation length of stay are well documented, but the postdischarge services that may replace some inpatient treatment appear to be greater than previously reported. The availability and impact of postdischarge care should be studied in greater detail to capture the wide array of postdischarge services and outcomes. Key Words: Health services research; Rehabilitation; Spinal cord injuries by the American Congress of Rehabilitation Medicine WHILE THERE IS CLEAR evidence that the average LOS for inpatient rehabilitation after SCI is shorter than in the past, the status of outpatient rehabilitation service delivery has barely been studied. In the United States, LOS has dropped dramatically over the last half-century. Information in the NSCID allows a consistent comparison over a 35-year period. For patients admitted to one of the Model Systems immediately after injury (ie, excluding transfers), the rehabilitation LOS dropped from a median of 98 days during 1973 to 1979, to 37 days during 2005 to The decline occurred in all SCI impairment groups, although not to the same degree (fig 1). 1 Managed care 2,3 and improvements in medical and rehabilitative expertise 4 are believed to account for declines in rehabilitation LOS 5,6 ; some worry that the reductions have gone too far. 3,7-9 LOS reductions may be associated with more AIS IRF LOS NSCID OT PBE PSY PT SCI SLP SW/CM TR VOC List of Abbreviations American Spinal Injury Association Impairment Scale inpatient rehabilitation facility length of stay National Spinal Cord Injury Database occupational therapy practice-based evidence psychology physical therapy spinal cord injury speech-language pathology social work/case management therapeutic recreation vocational counseling

2 362 INPATIENT AND POSTDISCHARGE SERVICES, Whiteneck Fig 1. Median days in rehabilitation unit by year of injury and neurologic category (patients admitted to the Model System immediately after injury only). Based on data from National Spinal Cord Injury Statistical Center. 1 efficient or more intense treatment, or simply less treatment, but one possibility is that outpatient care substitutes for inpatient treatment. Factors influencing the potential substitution of postdischarge for inpatient therapies might include a possible cost savings from reducing the service intensity provided in the hospital setting, better meeting the expectations of payers for shorter LOS, and better targeting of therapies to communityexperienced needs. However, the potential substitution of postdischarge for inpatient services also may increase the need for attendant care and the burden on families. Whether the reduction of inpatient SCI rehabilitation LOS may have been accompanied by a change in the amount of postdischarge services has been studied sparsely if at all. The SCI literature has little to say about the nature and extent of postdischarge rehabilitation services. While there are myriad articles describing specific interventions implemented as part of an outpatient or day rehabilitation program and the outcomes achieved, there is almost nothing published as to the specific rehabilitative services received after discharge from inpatient rehabilitation and for how long. 10,11 The only study found to report on outpatient therapy hours is that by Kogos et al, 12 who analyzed NSCID information for discharges between 1995 and While the average LOS in this period was fairly stable (mean, 52d), the percentage of patients who reported receiving any postdischarge PT, OT, or both, increased steadily from 24% to 74%. The percentage receiving at least 100 hours of OT and PT combined during the same time rose from 3% to 24%. The authors suggested that these increases were in reaction to LOS decreases in prior years. They did not find associations between the amount of postdischarge OT/PT and functional status, community integration, or quality of life at the first anniversary of injury, suggesting that the relationships between LOS, need for and receipt of postdischarge services, and shortterm outcomes are complex. Our own analysis of the NSCID data found that of 4071 patients interviewed in the years 2001 through 2003, around the first anniversary of their injury, 70.5% reported that postdischarge OT/PT services had been prescribed. Among those who received such services (with or without a prescription from the attending physiatrist) before the anniversary of their injury (67.7% of all patients), the average SD number of hours was hours. The percentage of patients who reported that postdischarge PSY, VOC, or both had been prescribed was 23.4%, and the average number of hours among those who received any of these services (21.7%) was only hours. OT/PT and PSY/VOC were received within the Model Systems outpatient programs and elsewhere. Unfortunately, the NSCID questions on postdischarge services were discontinued in 2006, so the link between changes in receipt and number of outpatient services and the continued decline in LOS cannot be analyzed for years after The goal of this study was to fill this gap in knowledge about the current status of SCI outpatient rehabilitation services, as well as inpatient services, by examining the relationships between therapeutic interventions provided during inpatient SCI rehabilitation and the delivery of rehabilitation services postdischarge from the inpatient rehabilitation setting. Specifically, the objective of the present analysis was to examine the variation in hours of therapy service by discipline, received in inpatient and postdischarge settings, during the first year after SCI. METHODS The SCIRehab Project is a 5-year multicenter study investigating details of the SCI rehabilitation process for nearly 1400 patients and relating them to first-year postinjury outcomes. The present study includes only patients enrolled in the first year of the SCIRehab study and analyzes a subset of the SCIRehab variables focusing on rehabilitation service delivery. Rehabilitation care provided during each patient encounter by clinicians involved in the inpatient rehabilitation process was documented in detail after each encounter on hand-held computers. Information on the nature and frequency of postdis-

3 INPATIENT AND POSTDISCHARGE SERVICES, Whiteneck 363 charge outpatient services received was collected directly from patients enrolled in the study via 2 postdischarge follow-up interviews. Rationale for and conceptualization of the SCIRehab project, including use of the PBE research methodology, have been described previously. 17,18 Discipline-specific taxonomies of treatment types and documentation systems were developed to capture details of daily treatments that are not contained in traditional documentation. These discipline-specific taxonomies for PT, OT, PSY, SW/CM, TR, SLP, and education and care coordination by nursing have been described SCIRehab Facilities and Enrollment Criteria The Rocky Mountain Regional Spinal Injury System at Craig Hospital leads the SCIRehab study and is joined by the Rehabilitation Institute of Chicago; Shepherd Center in Atlanta; National Rehabilitation Hospital in Washington, DC; Carolinas Rehabilitation in Charlotte, NC; and the Mount Sinai Medical Center in New York City. Willingness to participate in the study, geographic diversity, and treatment of large numbers of patients with SCI were used as selection criteria, and hence, the 6 centers are not a probability sample of rehabilitation facilities in the United States. These centers vary in care delivery patterns, patient clinical and demographic characteristics, and other factors that may affect outcomes. Institutional review board approval for this observational study was obtained at each SCIRehab facility, and all patients (or their parents or guardians) gave informed consent or assent. Patients who were 12 years or older and admitted for initial rehabilitation after traumatic SCI were enrolled in the project. Patients who received a substantial portion of their inpatient acute rehabilitation outside the participating center s SCI unit or program were excluded. A subset (493 patients) of the SCIRehab first-year sample of 600 patients was used for this analysis and included only patients who had completed their 12-month postdischarge interview. The 12-month interview was not able to be obtained for 6 patients who were deceased or incarcerated. Another 101 patients were lost to 12-month follow-up; however, 45 of these patients had completed their 6-month interview. A comparison of the 493 participants included in the study with the 107 excluded patients revealed significant differences in demographic characteristics, but no significant differences in injury characteristics. The excluded patients were more likely to be of minority race/ethnicity (43.9% vs 33.1%) and to not speak English (7.5% vs 1.4%), while less likely to be working at the time of injury (33.1% vs 43.9%) or to have education beyond high school (15.9% vs 29.4%). Data Sources Patient and injury data. The International Standards for Neurological Classification of SCI 26 were used to characterize the level and completeness of injury and to place study patients into 1 of 4 groups. Patients with AIS grade D (functionally motor incomplete) were grouped together regardless of injury level. For the other 3 groups, patients with AIS classifications of A (motor and sensory complete), B (sensory incomplete), and C (nonfunctional motor incomplete) were grouped together and then separated by motor neurologic level at rehabilitation admission: patients with cervical injury level 1 to 4 (high tetraplegia); those with cervical injury level 5 to 8 (low tetraplegia); and patients with paraplegia. A measure of the maximum overall medical acuity of the full rehabilitation stay for each participant (the maximum Comprehensive Severity Index) was calculated. 17 AIS classification, along with additional patient and injury data, was obtained through medical record review and from the NSCID maintained by the National Spinal Cord Injury Statistical Center in Birmingham, AL, which includes standardized information on injury-through-rehabilitation discharge (Form I) and postdischarge follow-up status (Form II). Length of stay. Rehabilitation LOS reflects the total time the patient spent in rehabilitation as calculated by subtracting the first admission date from the last discharge date and then subtracting days spent out of the rehabilitation facility (eg, in acute care or other interim setting) before returning for completion of rehabilitation. This definition and calculation was chosen so that the LOS would correspond to days on the SCI unit where service delivery was documented. Inpatient rehabilitation interventions. Treatment data from the SCIRehab project used in this analysis come from discipline-specific documentation that was completed by clinicians as they delivered care. Each discipline captured the number of minutes spent on each activity occurring within a treatment session, and these activity minutes were totaled to equal the approximate duration of the entire session. While all activities were documented for most disciplines, nursing interventions captured for the SCIRehab project only included direct patient education and case coordination; medicine, respiratory therapy, chaplaincy, and other minor therapies were not included. 17,18 Postdischarge rehabilitation interventions. An interviewer called each participant at 6 months and 12 months postinjury to complete an interview that was designed to obtain information about the person s health, functional and social status, quality of life, and utilization of postdischarge services. One of the questions asked was, Since your discharge from [inpatient] rehabilitation, have you received any of the following outpatient (or day program) rehabilitation services? At the 12- month interview, the question was, Since we last spoke on [date], have you received any of the following outpatient rehabilitation services? The interviewer then listed each service type (PT, OT, SLP, TR, PSY, nursing, activity-based training, SW/CM, and other ). For any service received, the respondent was then asked the approximate number of minutes in each treatment session, the number of sessions per week, and the number of weeks the service lasted. Data Processing and Analysis The total hours of inpatient rehabilitation treatment was calculated for each discipline by summing the durations of all sessions reported by that discipline between admission and discharge. Hours of inpatient rehabilitation treatment were added across reporting disciplines to estimate total inpatient treatment hours. Hours of postdischarge (outpatient or day hospital) treatment received from each discipline were calculated by multiplying session duration by the number of days per week times the number of weeks reported. Time spent on PT and activitybased training (activities focused on facilitating neural recovery and function below the level of injury, such as locomotor training and functional electrical stimulation cycling) was combined because the PT service typically provides activity-based training, and the two were included in the PT documentation for inpatient rehabilitation. The 6-month postinjury interview captured information about services received from the time of rehabilitation discharge to the date of the interview (which had a 6- to 8-month postinjury target window). The 12-month postinjury interview had a 12- to 18-month postinjury target window. To estimate the total amount of outpatient services received in the first year postinjury, the time covered by the

4 364 INPATIENT AND POSTDISCHARGE SERVICES, Whiteneck 12-month interview that occurred within the first year postinjury was estimated as a prorated proportion of all services reported in the 12-month interview, and then added to the service volume reported in the 6-month interview. The discipline-specific hours were added to obtain an estimate of the total hours of postdischarge services received before the first anniversary of injury. Descriptive statistics and box plots were used to characterize the amount and distribution of inpatient and postdischarge rehabilitation services received by study participants. Chisquare test (for nominal variables) and analysis of variance (for interval variables) were used to test for differences in patient and injury characteristics and the amount of services received among injury groups. Pearson correlation coefficients were calculated to assess the degree of association between inpatient and postdischarge services and among discipline treatment hours received postdischarge. Stepwise linear regression was used to identify predictors of the total hours of postdischarge service. RESULTS Patient and Injury Characteristics Table 1 presents patient and injury characteristics for the 493 patients and for each of 4 lesion level and impairment groups separately. Patients with high tetraplegia and AIS A, B, C comprised 21% of the sample and had a 76-day mean LOS; patients with low tetraplegia and AIS A, B, C were 26% of the sample and had a 67-day average LOS; those with paraplegia, AIS A, B, C were the largest group (38%) with a mean LOS of 45 days; and patients with AIS grade D made up 16% of the sample and had a 34-day mean LOS. Characteristics Table 1: SCIRehab Patient and Injury Characteristics Total Sample (N 493) C1-4 AIS A,B,C (n 102) C5-8 AIS A,B,C (n 126) Para AIS A,B,C (n 188) All AIS D (n 77) Inpatient rehabilitation LOS Time from injury to rehabilitation admission (d) Ventilator use at admission BMI at admission Admission motor FIM Admission cognitive FIM Maximum Comprehensive Severity Index Age at injury (y) Male Race/Ethnicity White Black Hispanic Other/unknown Work-related Injury Payer Medicare Medicaid Private insurance/payer Workers compensation Married at injury Education High school diploma High school diploma/ged High school diploma Other/unknown Employment status at injury Working Student Other Injury etiology Vehicular Violence Sports Fall or falling object Other/unknown NOTE. Values are mean SD, or percentages. Abbreviations: BMI, body mass index; FIM, Functional Independence Measure; GED, General Equivalency Diploma; Para, paraplegia.

5 INPATIENT AND POSTDISCHARGE SERVICES, Whiteneck 365 Hours of Inpatient and Postdischarge Rehabilitation Table 2 presents mean total hours of interventions during the inpatient rehabilitation stay and from the time of rehabilitation discharge to 1 year postinjury. The mean number of hours spent in inpatient and postdischarge therapy combined was 331 hours. A significant portion of rehabilitation hours, 44%, occurred after discharge from inpatient rehabilitation. Table 2 also provides the total hours of inpatient and postdischarge interventions by discipline, the percentage of patients with any postdischarge services, and the average time for those with any postdischarge services. Time spent on interventions during both the rehabilitation stay and the postdischarge period, and the proportion delivered after discharge varied by discipline. PT and OT provided the most therapy in both settings. Participants received 56% of their PT hours and 52% of their OT hours after discharge. This trend of receiving more PT and OT services postdischarge than during inpatient rehabilitation held for all injury categories except for the high tetraplegia group, which received more PT and OT as inpatients. Rehabilitation interventions provided after discharge by the other disciplines were much less extensive. While PT and OT provided postdischarge services to a majority of study participants, other disciplines served a minority. TR postdischarge services were received by 39.6%; PSY served 37.7%; SLP, 13.6%; SW/CM, 9.1%; and only 5.4% reported receiving outpatient nursing services. Variation in Inpatient and Postdischarge Rehabilitation Substantial variation was found in hours of total therapy received in both inpatient therapy and therapy provided postdischarge. The interquartile range was 134 ( ) hours for inpatient therapy and a much larger 190 (27 217) hours for postdischarge therapy. Greater variation was also found in the tails of the postdischarge therapy distribution compared with the inpatient distribution. While 5% of the patients received less than 47 hours of inpatient therapy and 5% received more than 404 inpatient hours, 8% of the patients received no postdischarge therapy, and 5% received more than 416 postdischarge hours. Figure 2 contrasts the variation in inpatient Services Table 2: SCIRehab Hours of Inpatient and Postdischarge Services Total Sample (N 434) C1-4 AIS A,B,C (n 93) C5-8 AIS A,B,C (n 112) Para AIS A,B,C (n 166) All AIS D (n 63) Inpatient Services (hours) All rehabilitation disciplines PT OT SLP TR PSY SW/CM Nursing Postdischarge Services (hours) All rehabilitation PT OT SLP TR PSY SW/CM Nursing Postdischarge Services (percentage of patients receiving any service; mean hours SD for patients with any services in the category) All rehabilitation disciplines PT OT SLP TR PSY SW/CM Nursing NOTE. Values are mean SD, or percentages. Abbreviation: Para, paraplegia.

6 366 INPATIENT AND POSTDISCHARGE SERVICES, Whiteneck Fig 2. Total hours of inpatient and postdischarge PT and OT during the first year postinjury. versus postdischarge PT and OT services, showing substantially greater variation in the hours of postdischarge services than in the hours of inpatient services. Figure 3 shows the variation in inpatient versus postdischarge service hours for the other disciplines, with the median (and sometimes the 75th percentile) equaling zero. Relationships Between Inpatient and Postdischarge Rehabilitation No significant association was found between the total hours of inpatient services and the total hours of postdischarge services received (r.08, P.085). The only disciplines with statistically significant correlations between inpatient and postdischarge service hours were OT (r.10, P.026) and TR (r.27, P.000). A strong correlation was found between the hours of PT and OT postdischarge treatment (r.80, P.001), indicating that the same patients who received a large amount of postdischarge PT also reported receiving more postdischarge OT. While weaker than the very strong PT/OT relationship, PT and OT postdischarge hours were significantly correlated with postdischarge TR hours (r.48 and.60, respectively) and with SLP hours (r.25 and.27, respectively), and TR hours were correlated with SLP hours (r.27), all at a P level of less than.001. The maximum (full-stay) Comprehensive Severity Index was the strongest predictor of the total hours of postdischarge services (R 2.04), with greater severity during rehabilitation associated with less postdischarge service. Of all the participant and injury characteristics reported in table 1, 11 variables were statistically significant in the stepwise regression model, but together they only accounted for 19% of the variance in the hours of postdischarge service. DISCUSSION This examination of inpatient and postdischarge rehabilitation services delivered within the first year after SCI revealed that a significant proportion of services are received after discharge from the inpatient setting. For both PT and OT, the total number of hours of postdischarge therapy was greater than the number of hours of inpatient therapy. Only a minority of patients received any postdischarge services from other disciplines, and the volume of postdischarge treatment was far less than that provided on an inpatient basis. While substantial variation was found in the total hours of inpatient services and in the hours delivered by each discipline, even greater variation was found in the distribution of postdischarge services. For both PT and OT, the interquartile range for postdischarge Fig 3. Total hours of inpatient and postdischarge services from other disciplines during the first year postinjury. Abbreviations: DC, discharge; RN, registered nurse.

7 INPATIENT AND POSTDISCHARGE SERVICES, Whiteneck 367 services was more than twice the interquartile range for inpatient services. A very high correlation was found between postdischarge hours of PT and OT treatment (and to a lesser degree TR and SLP hours), indicating that typically a combination of PT and OT (and sometimes TR and SLP) was delivered, with patients either receiving much or little service from these disciplines. This may indicate some facilities having multiple postdischarge services available and offered in a coordinated fashion. The very low correlation between total inpatient and postdischarge treatment hours indicates that there is neither a double benefit (or jeopardy), as might be indicated by a positive correlation (those receiving many inpatient hours also receive many outpatient hours), nor compensation, as might be indicated by a negative correlation (outpatient hours are increased for patients receiving few inpatient hours). A number of potential reasons could explain the partitioning of the total rehabilitation process between inpatient and postdischarge settings. Severity and level of injury, as well as comorbid conditions, commonly determine which individual rehabilitation goals can be accomplished in an inpatient setting and when a patient can be discharged from the inpatient setting to continue the rehabilitation process as an outpatient. Activity restrictions (imposed by postinjury or postsurgery precautions and complications or orthotic limitations) may influence treatment continuity and location, and dictate when a particular goal-directed therapy can be resumed, either as an inpatient or outpatient. Availability of expert postdischarge resources in a patient s home community may affect the timing of the inpatient-to-outpatient transition, particularly for patients living in rural areas. Funding source variations in payment policy (diagnosis-related group case payment, LOS limits, discounted fee for service, limits on postdischarge services, etc) often have an impact on the length of inpatient stay and the number of available outpatient treatment sessions. While none of the 4 payer groups entered the regression model as significant predictors of the hours of postdischarge services, more detailed information about variations in payer policies might reveal stronger relationships. The funding source issue is quite complex and also may interact with the clinician s ability to maximize treatment benefits for the individual patient. A review of the literature suggests that there is very limited health services research information available on the extent of outpatient rehabilitative services in general, or after a stay in an IRF or long-term care hospital, specifically. A substantial growth in the number of comprehensive outpatient rehabilitation facilities, from 72 to 522 between 1985 and 2000, has been reported, 27 which parallels the decline in LOS during that period for SCI rehabilitation as well as for rehabilitation for other conditions. 28 That trend has continued in the years since 2000, in rehabilitation LOS for SCI and other impairments, 29,30 but there is limited information on the growth of outpatient facilities or services. A report on postacute care services for the Medicare population noted that of those discharged in 2005 from an acute care hospital to an IRF (about 6% of all acute hospital discharges), 51.0% continued to use home health services after IRF discharge, 26.5% continued with outpatient therapy, and only 22.5% did not get further postacute care services. The numbers for 2006 were similar. 31 The growth in post-irf discharge rehabilitation services may have resulted in unexpectedly large utilization. In their conclusions on another PBE study, on joint replacement, DeJong et al 32 noted, Overall... the study uncovered much more postdischarge use of rehabilitation services than had been anticipated at the outset of the study. Patients with knee replacement received about as much therapy during the follow-up period as they did during the initial postacute setting [ie, inpatient rehabilitation or skilled nursing facility]. Patients with hip replacement received slightly more than half their PT and OT in an outpatient setting or from home health services. The amount and nature of postdischarge rehabilitation services provided are likely to differ among diagnostic groups because of differences in medical and rehabilitative needs, payer dynamics, and other factors. However, substantial use of postdischarge services is being reported in multiple populations. Another potential reason for an increase in postdischarge PT and OT SCI services is that many activity-based programs have been developed in response to the demand for the opportunity to maximize function. Study Limitations Several study limitations should be noted. By necessity, the method of collecting data on inpatient and postdischarge services varied. Inpatient data were collected prospectively by the treating clinicians, whereas postdischarge service data were collected retrospectively, using self-report from the person with SCI. Thus, the precision of the number of hours of postdischarge services by discipline should be kept in mind when comparing the inpatient with the postdischarge data. Postdischarge hours may be either underestimated or overestimated if a person was confused as to whether a therapist was, for example, a physical or an occupational therapist. Recall errors at 12 months about what happened before or after the prior interview might also be a problem, although interviewers had a copy of the 6-month responses to help minimize those problems. Investigating the reliability and validity of selfreported postdischarge rehabilitation service data should be a priority of future research. In addition to questions about postdischarge data accuracy, this report covers only the first year after SCI. Some patients may continue to receive outpatient therapy as an extension of what they received in the first year. Because postdischarge services are likely to diminish over time, the method of prorating service hours to only include the first year may have resulted in an underestimate. The average 12-month interview was conducted 14.6 months postinjury, and an average of 29% of services reported between the 6- and 12-month interviews were excluded from the first year postinjury analysis. The relatively high rehabilitation LOS (compared with the NSCID average) is another factor that may contribute to an overestimate of inpatient services and potentially an underestimate of postdischarge services. However, these limitations are mitigated by the robustness of the data. Substantial inaccuracies in the data would need to be present to change the major findings. CONCLUSIONS This study describes variations in inpatient and postdischarge services within the first year postinjury for persons undergoing rehabilitation for traumatic SCI at 6 centers participating in the SCIRehab study. It demonstrates that SCI rehabilitation care, at least for many patients, is provided in a continuum, with inpatient rehabilitation being only the beginning. Reductions in inpatient SCI rehabilitation LOS are well documented, but the postdischarge services that may replace some inpatient treatment appear to be greater than previously reported. Future studies should examine the availability and impact of postdischarge care in greater detail. These findings also emphasize the importance of considering the effects of both inpatient and postdischarge services when examining long-term outcomes of rehabilitative services.

8 368 INPATIENT AND POSTDISCHARGE SERVICES, Whiteneck References 1. National Spinal Cord Injury Statistical Center. Annual report for the Spinal Cord Injury Model Systems 2008 [table 51]. Birmingham: University of Alabama; Becker B, DeLisa J. Model Spinal Cord Injury System trends and implications for the future. Arch Phys Med Rehabil 1999;80: DeLisa J, Kirshblum S. A review: frustrations and needs in clinical care of spinal cord injury patients. J Spinal Cord Med 1997; 20: Ragnarsson K. Restorative treatment of persons with spinal cord injury: current trends. J Rehabil Res Dev 1998;35:xi-xiv. 5. Eastwood EA, Hagglund K, Ragnarsson K, Gordon W, Marino R. Medical rehabilitation length of stay and outcomes for persons with traumatic spinal cord injury Arch Phys Med Rehabil 1999;80: Fiedler I, Laud P, Maiman D, Apple D. Economics of managed care in spinal cord injury. Arch Phys Med Rehabil 1999;80: Meyers A, Bisbee A, Winter M. The Boston model of managed care and spinal cord injury: a cross-sectional study of the outcomes of risk-based, prepaid, managed care. Arch Phys Med Rehabil 1999;80: Ditunno J. Functional outcomes in spinal cord injury (SCI): quality care versus cost containment. J Spinal Cord Med 1997;20: Sipski M. Managed care and SCI. J Spinal Cord Med 1999;22: Riis V, Verrier M. Outpatient spinal cord injury rehabilitation: managing costs and funding in a changing health care environment. Disabil Rehabil 2007;29: Bloemen-Vrencken J, de Witte L, Post M. Follow-up care for persons with spinal cord injury living in the community: a systematic review of interventions and their evaluation. Spinal Cord 2005;43: Kogos SJ, DeVivo M, Richards J. Recent trends in spinal cord injury rehabilitation practices and outcomes. Top Spinal Cord Inj Rehabil 2004;10: Horn S, Gassaway J. Practice-based evidence study design for comparative effectiveness research. Med Care 2007;45(Suppl 2): S DeJong G, Horn S, Gassaway J, Stam HJ. Practice-based evidence for post-acute policy and practice: the case of joint replacement rehabilitation. Presented at Annual Conference of the American Congress of Rehabilitation Medicine; October 4-7, 2007; Washington, DC. 15. Gassaway J, Horn S, DeJong G, Smout R, Clark C, James R. Applying the clinical practice improvement approach to stroke rehabilitation: methods used and baseline results. Arch Phys Med Rehabil 2005;86(12 Suppl 2):S Horn S, DeJong G, Ryser D, Veazie P, Teraoka J. Another look at observational studies in rehabilitation research: going beyond the holy grail of the randomized controlled trial. Arch Phys Med Rehabil 2005;86(12 Suppl 2):S Whiteneck G, Dijkers M, Gassaway J, Jha A. SCIRehab: a new approach to study the content and outcomes of spinal cord injury rehabilitation. J Spinal Cord Med 2009;32: Gassaway J, Whiteneck G, Dijkers M. SCIRehab: clinical taxonomy development and application in spinal cord injury rehabilitation research. J Spinal Cord Med 2009;32: Natale A, Taylor S, LaBarbera J, et al. SCIRehab: the physical therapy taxonomy. J Spinal Cord Med 2009;32: Johnson K, Bailey J, Rundquist J, et al. SCIRehab: the supplemental nursing taxonomy. J Spinal Cord Med 2009;32: Ozelie R, Sipple C, Foy T, et al. SCIRehab: the occupational therapy taxonomy. J Spinal Cord Med 2009;32: Wilson C, Huston T, Koval J, Gordon S, Schwebel A, Gassaway J. SCIRehab: the psychology taxonomy. J Spinal Cord Med 2009; 32: Gordan W, Dale B, Brougham R, et al. SCIRehab: the speech language pathology taxonomy. J Spinal Cord Med 2009;32: Cahow C, Skolnick S, Joyce J, Jug J, Dragon C, Gassaway J. SCIRehab: the therapeutic recreation taxonomy. J Spinal Cord Med 2009;32: Abeyta N, Freeman E, Primack D, et al. SCIRehab: the social work/case management taxonomy. J Spinal Cord Med 2009;32: Marino R, editor. Reference manual for the International Standards for Neurological Classification of SCI. Chicago: American Spinal Injury Association; Bernstein A, Hing E, Moss A, Allen K, Siller A, Tiggle R, editors. Health care in America: trends in utilization. Hyattville: National Center for Health Statistics; Ottenbacher KJ, Smith PM, Illig SB, Linn RT, Ostir GV, Granger CV. Trends in length of stay, living setting, functional outcome, and mortality following medical rehabilitation. JAMA 2004;292: Dobrez D, Heinemann A, Deutsch A, Manheim L, Mallinson T. Impact of Medicare s prospective payment system for inpatient rehabilitation facilities on stroke patient outcomes. Am J Phys Med Rehabil 2010;89: Granger C, Markello S, Graham J, Deutsch A, Ottenbacher K. The Uniform Data System for medical rehabilitation: report of patients with stroke discharged from comprehensive medical programs in Am J Phys Med Rehabil 2009;88: Gage B, Morley M, Spain P, Ingber M, Waltham M. Examining post acute care relationships in an integrated hospital system [final report] Available at: pacihs/report.pdf. Accessed December 18, DeJong G, Tian W, Smout R, et al. Use of rehabilitation and other health care services by patients with joint replacement after discharge from skilled nursing facilities and inpatient rehabilitation facilities. Arch Phys Med Rehabil 2009;90: