Cognitive impairment predicts engagement in inpatient stroke rehabilitation : International Journal of Rehabilitation Research

Journal Logo

Original articles

Cognitive impairment predicts engagement in inpatient stroke rehabilitation

Lowder, Ryan J.a; Jaywant, Abhisheka,b,c; Fridman, Chaya B.b,c; Toglia, Joana,c,d; O’Dell, Michael W.a,c

Author Information
International Journal of Rehabilitation Research: December 2022 - Volume 45 - Issue 4 - p 359-365
doi: 10.1097/MRR.0000000000000552
  • Free

Abstract

Introduction

Patient engagement is an important component of inpatient stroke rehabilitation (ISR). Patient engagement refers to ‘an interest in, and an intentional effort to, work toward rehabilitation goals’ [1]. Patient engagement in ISR positively influences rehabilitation outcomes, with better engagement associated with shorter lengths of stay and greater improvement in functional status [2]. Participation and engagement during rehabilitation are also associated with improved outcomes in functional status following acute inpatient rehabilitation in combined samples of neurological and orthopedic illnesses [1,3–6].

Cognitive dysfunction may negatively impact patient engagement in ISR. Two studies have investigated the relationship between cognitive functioning and patient engagement in ISR using the Pittsburgh Rehabilitation Participation Scale (PRPS) [7]. These studies found that better executive functioning and visuospatial skills predicted increased engagement, whereas attention and memory did not [5,8]. However, these findings are limited by the choice of assessment used to measure engagement, as the PRPS comprises a single clinician-rated item. Though a single construct, engagement comprises multiple elements of motivation and participation that cannot be evaluated by one item [9]. Moreover, the generalizability of these initial studies was limited by the study sample, which included elderly subjects enrolled in an intervention trial for cognitive impairment. Given that many patients following stroke have mild and underrecognized cognitive impairment [10], investigating these questions in a broader sample may better delineate the association between cognition and engagement during ISR.

While various measures are used for cognitive assessment in ISR, a working group of experts from the National Institute of Neurological Disorders and Stroke (NINDS) and the Canadian Stroke Network (CSN) developed a neuropsychological screening battery to measure domain-specific cognitive impairment post-stroke. The 30-min version of the NINDS-CSN battery assesses processing speed, executive functioning, verbal memory, depression and other neuropsychiatric symptoms [11]. In a previous study, we demonstrated the sensitivity and clinical utility of the cognitive tests in this 30-min NINDS-CSN battery in ISR. Specifically, the cognitive tests of the battery showed sensitivity to cognitive impairment and had moderately strong internal consistency [12]. In the same study, the NINDS-CSN cognitive tests predicted functional gains during ISR and functional status at discharge. However, it is not yet known whether performance on the 30-min NINDS-CSN cognitive tests at admission is associated with the perception of rehabilitation engagement during inpatient rehabilitation.

The goal of the present study was to investigate the relationship between cognitive functioning on inpatient rehabilitation admission and clinician-rated rehabilitation engagement in patients with acute stroke. Cognitive functioning was assessed using a modified version of the 30-min NINDS-CSN cognitive battery. Specifically, we administered a modified version of the recommended NINDS-CSN protocol that (1) substituted the Symbol Digit Modalities Test (SDMT) for the Digit-Symbol Coding Test, as the SDMT includes the possibility of oral responding for participants with motor impairment and (2) we did not administer the Center for Epidemiological Studies-Depression subscale or the Neuropsychiatric Inventory because we chose to focus on cognition and minimize participant and staff burden within the context of a clinical inpatient unit. We refer to this modified assessment as the ‘30-min NINDS-CSN cognitive battery’ throughout to underscore the focus on the cognitive tasks. The engagement was assessed using the multidimensional Hopkins Rehabilitation Engagement Rating Scale (HRERS; [1]). Compared to the PRPS, the HRERS measures multiple elements of rehabilitation engagement, allowing us to assess the construct of engagement with greater reliability and comprehensiveness. We hypothesized that better performance on the 30-min NINDS-CSN cognitive battery would be associated with greater engagement in rehabilitation. Specifically, we hypothesized that NINDS-CSN subtests measuring executive function would be most predictive of patient engagement based on existing findings in the engagement literature [5,8,12,13].

Methods

Participants

Individuals who participated in this study were a subset of a larger cohort of patients with mild to moderate stroke severity admitted to ISR at an urban academic medical center between August 2012 and July 2016. The study was approved by the Weill Cornell Medicine Institutional Review Board. Of the 423 patients admitted to ISR during the period of study enrollment, 273 inpatients met inclusion criteria and provided informed consent for their routine clinical data to be stored and utilized for research. Study inclusion criteria for the larger cohort included admission to ISR, radiographic diagnosis of stroke, age ≥18 years, language expression and comprehension sufficient to complete the cognitive tasks and English proficiency. In the present study, 110 patients who had the HRERS completed by a treating therapist and had each completed the majority of the NINDS-CSN cognitive battery were included for analysis.

Measures

30-min NINDS-CSN cognitive battery

The cognitive battery is comprised of neuropsychological tests that include animal naming, the Controlled Oral Word Association Test (COWAT; letters C, F, and L), Wechsler Adult Intelligence Scale – Fourth Edition (WAIS-IV) Coding subtest, Trail Making Test A and B (TMT A & B) and Hopkins Verbal Learning Test – Revised (HVLT-R) Total Recall, Delayed Recall and Recognition subtests [11,14–17]. In anticipation of poststroke motor impairments present in the study sample, the WAIS-IV Coding subtest, which can only be administered in written format, was substituted with the SDMT, another measure of processing speed that has the option for oral responding and that has been validated in patients with stroke [18]. Both the written and oral formats of the SDMT were utilized, depending on the patients’ motor status. The 30-min NINDS-CSN cognitive battery was administered within 72 h of admission to the inpatient rehabilitation unit. A speech-language pathologist administered the Animal Naming, COWAT and HVLT-R subtests and an occupational therapist administered the SDMT and TMT subtests. The therapists were blind to planned analyses.

Each 30-min NINDS-CSN cognitive battery subtest raw score was converted into a standard (T) score based on previously published normative data. The COWAT, TMT and HVLT-R subtest scores were corrected for age, and the Animal Naming and SDMT subtest scores were corrected for age and education [19–24]. The 30-min NINDS-CSN cognitive battery total score was calculated by averaging the T scores for each of the battery’s subtests, following prior research on the battery [12,25].

Hopkins Rehabilitation Engagement Rating Scale

The HRERS is a 5-item therapist rating scale of patient engagement during rehabilitation therapy. Items include assessment of therapy attendance, prompts required during therapy, attitude towards therapy, acknowledgment of therapy need and active participation during therapy. Each item is measured using a 6-point Likert scale, from ‘Never’ to ‘Always’. HRERS scores range from 0 to 30, with a higher score indicating greater engagement. The HRERS represents a unidimensional construct that measures multiple elements of rehabilitation engagement, has an established criterion validity, and has high internal consistency and good interrater reliability [1]. The HRERS was completed by a treating occupational therapist upon patient discharge to measure the patient’s engagement during their entire inpatient rehabilitation admission. The rating therapists were blind to planned analyses.

2-item Patient Health Questionnaire

The 2-item Patient Health Questionnaire (PHQ-2) was used to screen for depression in the inpatient rehabilitation unit [26]. In total 88/110 individuals in our sample had PHQ-2 scores available in the medical record. Individuals who endorsed either depressed mood or loss of interest on the PHQ-2 were considered to be a positive screen for depression and were subsequently administered the 9-item PHQ. Given that only 11 participants in our sample screened positive for depression, we elected to use the PHQ-2 as a categorical variable (positive screen for depression, negative screen for depression) to explore the association between depression and engagement.

Statistical analyses

All analyses were conducted using IBM SPSS version 27. Past research on the HRERS found a scoring ceiling effect [1], with the majority of the study sample receiving perfect or near-perfect scores. The developers of the HRERS suggested three clinically different groups of scores, with scores >25 considered ‘normal’, scores 20–25 considered ‘at risk’ for less engagement and poorer outcomes, and scores <20 considered ‘in need of intervention’ to avoid less engagement and poorer outcomes [1]. In the current study, a disproportionate number of participants scored >25 so we divided participants into two groups: a ‘high’ engagement group (with scores >25) and a ‘low’ engagement group (combining the ‘at risk’ and ‘in need of intervention’ groups, with scores ≤25).

Group differences between high and low engagement groups were compared using independent-sample t-tests for normally distributed variables, as confirmed by Kruskal Wallis tests, and Mann-Whitney U tests for non-normally distributed variables. Parametric and nonparametric Cohen’s d’s were calculated for effect sizes of group differences [27]. Group differences between engagement groups for categorical variables were assessed using chi-square tests or Fisher’s exact tests depending on cell count. Spearman correlations were conducted to evaluate the relationships between scores on the 30-min NINDS-CSN cognitive battery total and subtests (using T scores as continuous variables) and the HRERS. Statistical significance for Spearman correlations was evaluated using a Bonferroni-corrected P value of 0.006.

High and low engagement groups were used as dependent outcomes for Firth logistic regression models controlling for age, stroke severity and prior stroke. Firth logistic regression models were calculated for NINDS-CSN cognitive subtests that significantly correlated with HRERS scores. Separate models were used for the battery total and subtests to avoid multicollinearity. Firth logistic regressions, which utilize penalized maximum-likelihood estimations, were used to reduce bias introduced by small sample sizes and uneven distributions between dichotomous outcomes [28]. Firth logistic regressions also minimize the error associated with having less than 10 events per variable, mainly the risk of separation, permitting more covariates to be included in each model [29].

Finally, to explore the association between depression and engagement, we conducted a Fisher’s exact test using a 2 × 2 contingency table of engagement group (high vs. low) × depression status (positive screen vs. negative screen).

Results

Study sample demographics and clinical characteristics

Clinical and demographic comparisons between the 110 participants included in the high (N = 76) versus low (N = 34) engagement groups are presented in Table 1. Participants in the low engagement group were older (P = 0.013), had greater stroke severity (P = 0.017), had more functional impairments at ISR admission (P < 0.0001), and were less likely to be admitted for a first-time stroke (P = 0.029) compared to the high engagement group.

Table 1 - Sample demographics and characteristics
High engagement (N = 76) Low engagement (N = 34) P value
Age in years median (range)a 69 (27–92) 79 (26–98) 0.013
Gender 0.952
 Male 34 (44.7%) 15 (44.1%)
 Female 42 (55.3%) 19 (55.9%)
Education 0.838
 Less than high school 5 (6.6%) 3 (8.8%)
 Completed high school 16 (21.1%) 10 (29.4%)
 Some college 11 (14.5%) 4 (11.8%)
 Completed college 28 (36.8%) 12 (35.3%)
 Graduate degree 16 (21.1%) 5 (14.7%)
Handedness 0.876
 Right 67 (88.2%) 30 (88.2%)
 Left 7 (9.2%) 4 (11.8%)
 Ambidextrous 2 (2.6%) 0 (0.0%)
Lesion side 0.427
 Left 37 (48.7%) 21 (61.8%)
 Right 31 (40.8%) 10 (29.4%)
 Bilateral 8 (10.5%) 3 (8.8%)
First time stroke 0.029
 Yes 60 (78.9%) 20 (58.8%)
 No 16 (21.1%) 14 (41.2%)
NIHSS median (range)a 4.0 (0–31) 6.0 (0–25) 0.017
Baseline FIM total 70.7 ± 15.2 57.1 ± 19.9 <0.0001
CCI median (range)a 1 (0–8) 1 (0–5) 0.896
Values are mean ± standard deviation or frequency (percentage), unless otherwise indicated.
CCI, Charlson Comorbidity Index; FIM, functional independence measure; NIHSS, National Institute of Health Stroke Scale.
aFor ordinal variables and variables with non-normal distributions, nonparametric tests were conducted.

Group differences in cognitive functioning

We found that patients within the high-engagement group performed significantly better than participants in the low-engagement group on the 30-min NINDS-CSN cognitive battery total, SDMT, TMT A, Animal Naming and HLVT-Recognition (all P’s < 0.05). Differences between scores on the TMT B, COWAT and HVLT-R Total Recall trended towards significance (all P’s < 0.10) and scores on the HVLT-R Delayed Recall did not differ between groups (Table 2).

Table 2 - NINDS-CSN cognitive battery performance between high and low engagement groups
NINDS-CSN cognitive battery total and subtests High engagement Low engagement Test statistic (t or U) Effect size P value
Mean or median SD or IQR Mean or median SD or IQR
NINDS-CSN Totala 34.4 16.0 27.8 11.0 1776.5 0.6 0.002
SDMTb 32.9 10.0 25.0 9.1 −3.7 0.8 <0.0001
TMT-Aa 31.0 36.0 14.0 14.0 1437.5 0.6 0.003
TMT-Ba 32.0 26.3 24.0 18.0 1037.0 0.4 0.056
Animal Naminga 37.5 16.8 31.0 11.0 1340.5 0.6 0.005
COWATa 37.0 20.0 35.0 12.5 1176.5 0.4 0.062
HVLT-R Total Recallb 33.5 12.0 29.5 9.9 −1.7 0.4 0.098
HVLT-R Delayed Recalla 33.0 24.5 26.0 16.0 1135.0 0.2 0.371
HVLT-R Recognitiona 39.5 34.8 30.0 28.0 1511.0 0.4 0.035
COWAT, Controlled Oral Word Associated Test; HVLT-R, Hopkins Verbal Learning Test-Revised; IQR, interquartile range; NIHSS, National Institute of Health Stroke Scale; NINDS-CSN, National Institute of Neurological Disorders and Stroke-Canadian Stroke Network; SDMT, Symbol Digit Modalities Test; TMT, Trail Making Test.
aFor variables with non-normal distributions, nonparametric tests were conducted and median interquartile range (IQR) are reported.
bMean and SD are reported for normally distributed variables.

Correlation between cognitive functioning and rehabilitation engagement

Scores on the 30-min NINDS-CSN cognitive battery total, SDMT, TMT A, Animal Naming and COWAT correlated significantly with raw HRERS scores using a Bonferroni-corrected P value (all rs > 0.292, all P’s < 0.006) (Table 3). These variables were entered as predictor variables in separate Firth logistic regressions, with the engagement group as the outcome variable.

Table 3 - Spearman correlations results
Neuropsychological tests r s
NINDS-CSN cognitive battery total 0.37a
SDMT 0.46a
TMT-A 0.37a
Animal Naming 0.32a
COWAT 0.29a
TMT-B 0.28b
HVLT-R Recognition 0.23b
HVLT-R Total Recall 0.20b
HVLT-R Delayed Recall 0.11
COWAT, Controlled Oral Word Associated Test; HVLT-R, Hopkins Verbal Learning Test-Revised; NIHSS, National Institute of Health Stroke Scale; SDMT, Symbol Digit Modalities Test; TMT, Trail Making Test.
aP < 0.006 (Bonferroni-corrected P value).
bP < 0.05.

Results of the Firth logistic regressions, each controlling for age, stroke severity and prior stroke, indicated that higher scores on the NINDS-CSN cognitive battery total, SDMT and TMT-A significantly increased the odds of being categorized as a ‘high engager’ (all P’s < 0.05) (Table 4). The COWAT and Animal Naming subtests were not associated with engagement after adjusting for age, stroke severity and prior stroke.

Table 4 - Firth logistic regression models
NINDS-CSN subtests and covariates Coefficient OR SE FPML χ2 Pvalue 95% CIs
Model 1 <0.0001
 NINDS-CSN cognitive battery total 0.07 1.07 0.03 7.59 0.006 1.02–1.13
 Age −0.04 0.96 0.02 5.67 0.017 0.93–0.99
 NIHSS −0.05 0.95 0.04 1.88 0.170 0.88–1.02
 Prior stroke −0.79 0.45 0.49 2.63 0.105 0.17–1.18
Model 2 0.002
 SDMT 0.07 1.07 0.02 7.98 0.005 1.02–1.13
 Age −0.02 0.98 0.02 2.08 0.150 0.94–1.01
 NIHSS −0.05 0.95 0.04 1.79 0.181 0.88–1.03
 Prior stroke −0.75 0.47 0.53 2.05 0.152 0.17–1.32
Model 3 0.002
 TMT-A 0.04 1.04 0.02 5.30 0.021 1.01–1.07
 Age −0.03 0.97 0.02 4.00 0.046 0.93–1.00
 NIHSS −0.05 0.95 0.04 1.76 0.185 0.88–1.02
 Prior stroke −0.79 0.45 0.51 2.50 0.114 0.17–1.21
Model 4 0.003
 Animal naming 0.04 1.04 0.02 3.75 0.053 1.00–1.09
 Age −0.02 0.98 0.02 2.21 0.137 0.95–1.01
 NIHSS −0.05 0.97 0.04 2.05 0.152 0.88–1.02
 Prior stroke −1.06 0.35 0.52 4.42 0.035 0.13–0.93
Model 5 0.010
 COWAT 0.03 1.02 0.02 1.35 0.245 0.98–1.07
 Age −0.03 0.97 0.02 3.01 0.083 0.94–1.00
 NIHSS −0.06 0.94 0.04 2.65 0.103 0.87–1.01
 Prior stroke −01.05 0.35 0.55 3.81 0.051 0.12–1.00
Models were only calculated for NINDS-CSN cognitive battery subtests that correlated with raw HRERS scores. Coefficient = Firth Logistic Regression Coefficient. Total score and individual subtests were entered in separate models adjusting for stroke severity.
COWAT, Controlled Oral Word Association Test; FPML, Firth Penalized Maximum Likelihood; NIHSS, National Institute of Health Stroke Scale; NINDS-CSN, National Institute of Neurological Disorders and Stroke-Canadian Stroke Network; SDMT, Symbol Digit Modalities Test; TMT, Trail Making Test.

Association between depression and engagement

Eleven out of 88 participants who had PHQ-2 scores screened positive for depression (12.5%). Results of a Fisher’s exact test did not indicate a statistically significant difference in positive vs. negative screen for depression and engagement group (high vs. low; P = 0.72).

Discussion

The purpose of this study was (1) to evaluate the relationship between performance on a standardized cognitive battery and patient engagement in ISR in a broad sample of patients with stroke and a comprehensive assessment of engagement; and (2) to determine whether specific cognitive abilities could predict high versus low engagement in ISR.

As expected, we found that participants in the high-engagement group scored significantly better on a modified version of the 30-min NINDS-CSN cognitive battery with medium-to-large effect sizes compared to participants in the low-engagement group. Specifically, the SDMT and TMT subtests, which measure various components of executive functioning and attention, including visuospatial attention, working memory, task-switching and processing speed [30,31], showed the greatest differences between groups and strongest correlations with the HRERS. These subtests were also modest but significant predictors of engagement group in the regression analysis accounting for age, stroke severity, and prior stroke.

We used the COWAT and Animal Naming subtests to further assess executive functioning, as well as language processing [32,33]. Given that participants who may be expected to perform worse on language processing tasks, including patients without English proficiency or with clinically significant aphasia, were excluded, it is reasonable to consider performance on the verbal fluency assessments as largely indicative of executive functioning. The COWAT and Animal Naming subtests differed significantly between engagement groups and significantly correlated with HRERS. However, these tests were not significant predictors of engagement after correcting for covariates in the logistic regression analyses, though the Animal Naming subtest was trending towards significance.

Verbal memory was largely not associated with rehabilitation engagement. Performance on the HVLT-R Total Recall and Delayed Recall subtests did not differ significantly between engagement groups and were not correlated with the HRERS, which corresponds with previous research [5,8]. While the HVLT-R Recognition subtest showed group differences and correlated with the HRERS in a univariate correlation, it did not predict the engagement group after adjusting for covariates. The lack of association between verbal memory and engagement is generally consistent with our hypothesis and with prior findings that memory is not strongly linked to rehabilitation outcomes [12]. This is reasonably expected in the setting of ISR, where patients are completing activities with therapist support, so memory impairment may have less of an impact on functioning.

Given that the subtests that predicted patient engagement in our sample were all timed tasks, we suggest that deficits in attention and executive functioning may be, at least partially, secondary to reduced processing speed, which was not investigated in previous studies. Stroke patients with impaired processing speed may have difficulty with managing rapidly incoming information, thus making it difficult for them to keep up with the rigorous demands of inpatient rehabilitation and exacerbating the effects of dysfunction in other cognitive domains. Patients with poor executive functions may exhibit lack of initiation or face difficulty when required to switch between tasks, making it challenging for them to fully join in and participate in therapy. Patients with poor attention may struggle with focusing on a particular task during rehabilitation, which could also limit their ability to be fully engaged in therapy.

Rehabilitation clinicians may consider placing greater weight on the effects of cognitive deficits in rehabilitation engagement and long-term outcomes. In a study that surveyed therapists on their perceived biggest barrier to patient engagement during rehabilitation, attentional deficits and general cognitive impairment were reported as a barrier by only 7.1 and 8.3%, respectively, of therapists [34]. Identification of cognitive deficits through formal assessment may inform clinical decision-making in ISR by guiding interventions and minimizing potential biases and barriers to care.

Making modifications to therapy to accommodate executive functioning deficits, such as reducing environmental distractions, presenting simplified material, slowing pace of instruction and providing frequent guidance and positive feedback, may help improve engagement in therapy. Psychosocial, psychotherapeutic and recreation therapy programs may also enhance engagement in rehabilitation. While research supporting interventions for executive functioning and attention impairment remains limited [35], there is evidence for the feasibility and efficacy of metacognitive strategy training programs targeting executive functioning [12,36,37]. Researchers who study such interventions in ISR may consider including patient engagement as an outcome to further study the relationship between cognition and engagement.

While our results generally accord with prior research, we found subtle discrepancies between our findings and prior studies of the engagement-cognition relationship, the latter of which found performance on executive function and visuospatial assessments to be most predictive of rehabilitation engagement, compared to attention. These differences may reflect the choice of neuropsychological tests administered. Executive functions, which include inhibitory control, working memory, and cognitive flexibility, are often intertwined with attentional processes [38]. The present study also comprised a more generalizable sample of participants with stroke compared to previous studies, including patients who did not screen positive for cognitive impairment. For this reason, our data may better capture the full spectrum of cognitive function in stroke patients undergoing rehabilitation.

The present study further supports the clinical utility of a modified version of the 30-min NINDS-CSN cognitive battery by demonstrating its ability to predict rehabilitation engagement. The 30-min NINDS-CSN cognitive battery has been shown to be useful in identifying cognitive impairment that is missed by shorter screening measures and predicts rehabilitation gain and independence in instrumental activities of daily living at inpatient stroke rehabilitation discharge [12]. The 30-min NINDS-CSN cognitive battery is clinically feasible in inpatient and outpatient stroke settings and has moderate to excellent discriminatory power for identifying cognitive impairment [39,40]. Of specific subtests, the SDMT, which is not included in the original NINDS-CSN cognitive battery and was used in this study to accommodate for motor impairments, was most strongly correlated with and best predicted rehabilitation engagement. If neuropsychological assessment upon ISR admission is constrained, the SDMT may be the most useful subtest to gauge and intervene upon patient engagement.

Finally, in an exploratory analysis, we did not find an association between the result of a depression screen (positive or negative for depression) and the engagement group. Previous research has demonstrated an association between depression and engagement in inpatient rehabilitation [1]. Clinically, the presence of symptoms such as low mood, anhedonia/motivation, and low self-esteem can interfere with engagement in rehabilitation. We did not find an association between depression and engagement may reflect the relative insensitivity of the PHQ-2 to depression symptoms and their severity [41].

Study limitations

The participants who were included in the present analyses were limited to individuals with mild-moderate stroke and relatively high education levels. Thus, our findings are generalizable to relatively well-educated and English-speaking individuals with mild-moderate stroke severity. Furthermore, participants that choose to participate in research studies may be more likely to be more engaged in rehabilitation therapy, which could explain the disproportionate number of participants that were categorized as high engagers in this study. The NINDS-CSN cognitive battery does not include certain cognitive domains (visuospatial functioning and visual memory) and thus we were unable to evaluate the impact of those domains on engagement. Our screening measure for depression was relatively limited and future research should incorporate a robust depression symptom severity assessment to better understand the association between depression and engagement. We also did not have access to additional variables (e.g. medications) that may have impacted engagement. Finally, we did not measure patient self-reported perception of barriers to engagement, which is a relevant variable to inform interventions to improve engagement.

Conclusion

In summary, we demonstrated that a modified version of the NINDS-CSN cognitive battery and its subtests that measure processing speed, attention and executive function are predictive of patient engagement in ISR. Specifically, these findings suggest that impairments in these cognitive domains can provide clinicians with important information regarding engagement in rehabilitation therapy. Future research should explore whether there is a causal relationship between post-stroke cognitive impairment in these domains and rehabilitation engagement. Clinical trials investigating cognitive rehabilitation interventions may consider utilizing patient engagement as an outcome measure.

Acknowledgements

This study was supported by the Peter Jay Sharp Foundation.

ClinicalTrials.gov identifier NCT02876783.

Conflicts of interest

There are no conflicts of interest.

References

1. Kortte KB, Falk LD, Castillo RC, Johnson-Greene D, Wegener ST. The Hopkins rehabilitation engagement rating scale: development and psychometric properties. Arch Phys Med Rehabil. 2007; 88:877–884.
2. Dorra HH, Lenze EJ, Kim Y, Mulsant BH, Munin MC, Dew MA, et al. Clinically relevant behaviors in elderly hip fracture inpatients. Int J Psychiatry Med. 2002; 32:249–259.
3. Lenze EJ, Munin MC, Quear T, Dew MA, Rogers JC, Begley AE, et al. Significance of poor patient participation in physical and occupational therapy for functional outcome and length of stay. Arch Phys Med Rehabil. 2004; 85:1599–1601.
4. Munin MC, Seligman K, Dew MA, Quear T, Skidmore ER, Gruen G, et al. Effect of rehabilitation site on functional recovery after hip fracture. Arch Phys Med Rehabil. 2005; 86:367–372.
5. Skidmore ER, Whyte EM, Holm MB, Becker JT, Butters MA, Dew MA, et al. Cognitive and affective predictors of rehabilitation participation after stroke. Arch Phys Med Rehabil. 2010; 91:203–207.
6. Paolucci S, Di Vita A, Massicci R, Traballesi M, Bureca I, Matano A, et al. Impact of participation on rehabilitation results: a multivariate study. Eur J Phys Rehabil Med. 2012; 48:455–466.
7. Lenze EJ, Munin MC, Quear T, Dew MA, Rogers JC, Begley AE, et al. The Pittsburgh rehabilitation participation scale: reliability and validity of a clinician-rated measure of participation in acute rehabilitation. Arch Phys Med Rehabil. 2004; 85:380–384.
8. Kringle EA, Terhorst L, Butters MA, Skidmore ER. Clinical predictors of engagement in inpatient rehabilitation among stroke survivors with cognitive deficits: an exploratory study. J Int Neuropsychol Soc. 2018; 24:572–583.
9. Rettke H, Geschwindner HM, van den Heuvel WJA. Assessment of patient participation in physical rehabilitation activities: an integrative review. Rehabil Nurs. 2015; 40:209–223.
10. Wolf TJ, Baum C, Conner LT. Changing face of stroke: implications for occupational therapy practice. Am J Occup Ther. 2009; 63:621–625.
11. Hachinski V, Iadecola C, Petersen RC, Breteler MM, Nyenhuis DL, Black SE, et al. National institute of neurological disorders and stroke-Canadian stroke network vascular cognitive impairment harmonization standards. Stroke. 2006; 37:2220–2241.
12. Jaywant A, Toglia J, Gunning FM, O’Dell MW. The clinical utility of a 30-minute neuropsychological assessment battery in inpatient stroke rehabilitation. J Neurol Sci. 2018; 390:54–62.
13. Park YH, Jang J-W, Park SY, Wang MJ, Lim J-S, Baek MJ, et al. Executive function as a strong predictor of recovery from disability in patients with acute stroke: a preliminary study. J Stroke Cerebrovasc Dis. 2015; 24:554–561.
14. Strauss E, Sherman EMS, Spreen O. A compendium of neuropyschological tests. The Oxford University Press; 2006.
15. Smith A. Symbol digit modalities test. Western Psychological Services; 2017.
16. Reitan RM, Wolfson D. Category test and trail making test as measures of frontal lobe functions. Clin Neuropsychol. United Kingdom: Taylor & Francis. 1995; 9:50–56.
17. Brandt J, Benedict RHB. Hopkins verbal learning test-revised professional manual. PAR; 2001.
18. Koh C-L, Lu W-S, Chen H-C, Hsueh I-P, Hsieh J-J, Hsieh C-L. Test-retest reliability and practice effect of the oral-format symbol digit modalities test in patients with stroke. Arch Clin Neuropsychol. 2011; 26:356–363.
19. Ruff RM, Light RH, Parker SB, Levin HS. Benton controlled oral word association test: reliability and updated norms. Arch Clin Neuropsychol. 1996; 11:329–338.
20. Ivnik RJ, Malec JF, Smith GE, Tangalos EG, Petersen RC. Neuropsychological tests’ norms above age 55: COWAT, BNT, MAE token, WRAT-R reading, AMNART, STROOP, TMT, and JLO. Clin Neuropsychol Swets & Zeitlinger. 1996; 10:262–278.
21. Tombaugh TN, Kozak J, Rees L. Normative data stratified by age and education for two measures of verbal fluency: fas and animal naming. Arch Clin Neuropsychol. 1999; 14:167–177.
22. Tombaugh TN. Trail making test A and B: normative data stratified by age and education. Arch Clin Neuropsychol. 2004; 19:203–214.
23. Benedict RHB, Schretlen D, Groninger L, Brandt J. Hopkins verbal learning test – revised: normative data and analysis of inter-form and test-retest reliability. Clin Neuropsychol Routledge. 1998; 12:43–55.
24. Kiely KM, Butterworth P, Watson N, Wooden M. The symbol digit modalities test: normative data from a large nationally representative sample of Australians. Arch Clin Neuropsychol. 2014; 29:767–775.
25. Chen X, Wong A, Ye R, Xiao L, Wang Z, Lin Y, et al. Validation of NINDS-CSN neuropsychological battery for vascular cognitive impairment in Chinese stroke patients. BMC Neurol. 2015; 15:20.
26. Kroenke K, Spitzer RL, Williams JBW. The patient health questionnaire-2: validity of a two-item depression screener. Med Care. 2003; 41:1284–1292.
27. Fritz CO, Morris PE, Richler JJ. Effect size estimates: current use, calculations, and interpretation. J Exp Psychol Gen. 2012; 141:2–18.
28. Heinze G, Schemper M. A solution to the problem of separation in logistic regression. Stat Med. 2002; 21:2409–2419.
29. van Smeden M, de Groot JAH, Moons KGM, Collins GS, Altman DG, Eijkemans MJC, et al. No rationale for 1 variable per 10 events criterion for binary logistic regression analysis. BMC Med Res Methodol. 2016; 16:163.
30. Smith A. Symbol digit modalities test. Western Psychological Services; 1982.
31. Sánchez-Cubillo I, Periáñez JA, Adrover-Roig D, Rodríguez-Sánchez JM, Ríos-Lago M, Tirapu J, et al. Construct validity of the trail making test: role of task-switching, working memory, inhibition/interference control, and visuomotor abilities. J Int Neuropsychol Soc. 2009; 15:438–450.
32. Whiteside DM, Kealey T, Semla M, Luu H, Rice L, Basso MR, et al. Verbal fluency: language or executive function measure? Appl Neuropsychol Adult. 2016; 23:29–34.
33. Ruff RM, Light RH, Parker SB, Levin HS. The psychological construct of word fluency. Brain Lang. 1997; 57:394–405.
34. Lequerica AH, Donnell CS, Tate DG. Patient engagement in rehabilitation therapy: physical and occupational therapist impressions. Disabil Rehabil. 2009; 31:753–760.
35. Chung C, Pollock A, Campbell T, Durward B, Hagen S. Cognitive rehabilitation for executive dysfunction in adults with stroke or other adult nonprogressive acquired brain damage. Stroke. 2013; 44:e77–e78.
36. Toglia J, Johnston MV, Goverover Y, Dain B. A multicontext approach to promoting transfer of strategy use and self regulation after brain injury: an exploratory study. Brain Inj. 2010; 24:664–677.
37. Toglia J, Lee A, Steinberg C, Waldman-Levi A. Establishing and measuring treatment fidelity of a complex cognitive rehabilitation intervention: the multicontext approach. Br J Occup Ther SAGE Publications Ltd STM. 2020; 83:363–374.
38. Diamond A. Executive functions. Annu Rev Psychol. 2013; 64:135–168.
39. Han DY, Anderson AJ, Jones JE, Hermann BP, Sattin JA. Neuropsychology in multidisciplinary stroke care: clinical feasibility of the NINDS-CSN vascular cognitive impairment harmonization standards. Int Sch Res Notices. 2014; 2014:216024.
40. Xu X, Chan QL, Hilal S, Ikram MK, Venketasubramanian N, Tan BY, et al. The diagnostic utility of the NINDS-CSN neuropsychological battery in memory clinics. Dement Geriatr Cogn Dis Extra. 2016; 6:276–282.
41. Shankar L, Smith N, Uchino K, Thompson NR, Pozuelo L, Katzan IL. Evaluation of the patient health questionnaire-2 as a screening tool for depression during the acute stroke admission. J Stroke Cerebrovasc Dis. 2017; 26:2519–2526.
Keywords:

activities of daily living; cerebrovascular disease; motivation; neuropsychological assessment; stroke

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.