The DRESS trial: a feasibility randomized controlled trial of a neuropsychological approach to dressing therapy for stroke inpatients (2024)

Abstract

Introduction

Dressing is a daily activity which is taken for granted by the able-bodied. Followingstroke this self-care task can be problematic, with 54% of stroke survivors unable to dressindependently at six months¹ and 36% at two years after stroke.² The prevalence of this problem isunsurprising, however, as dressing is a complex skill that requires many physical andcognitive abilities to ensure independence.³ Previous longitudinal studies havedocumented that those patients with persistent cognitive difficulties have higher levels ofdressing dependence than those without² and that the nature of cognitivedifficulties determines the pattern of persistent dressing problems.⁴

As part of routine stroke rehabilitation, occupational therapists assess the self-careabilities of each patient and strive to resolve any dressing difficulties observed. Anarrative literature review⁵ and survey of occupational therapy dressing practices in the UK⁶ documented that therapistsdid not use standardized dressing assessments to evaluate dressing performance nor did theyuse research evidence to inform their clinical practice, frequently providing atime-limited, repetitive, problem-solving approach to dressing practice. This method isreferred to as the ‘functional approach’ to dressing. Although cognitive deficits wereacknowledged as a key prohibitive factor in the acquisition of dressing independence, therewas little evidence in the survey of therapists tailoring the approach to dressing treatmentin light of impairments experienced by the patient. Therapists who did treat cognitiveimpairments did so using mental stimulation on unrelated cognitive exercises, such as penand paper exercises to improve visual neglect, in the hope that improvements on these tasksmight generalize to dressing ability. There was little evidence of therapists treating thecognitive difficulties directly during dressing practice.

There is some evidence that dressing practice provided by occupational therapists can bebeneficial. A previous single-blind, randomized cross-over trial has described thesuccessful treatment of dressing difficulties after stroke in a group of community-dwellingpatients (n = 30) at six months after stroke.⁷ Dressing practice was administered by anexperienced occupational therapist over a three-month period and employed a pragmaticfunctional approach to treatment. The interventions provided included advice on appropriateclothing, the teaching of strategies such as dressing the affected side first, the use ofmarkers on garments to overcome perceptual problems and energy conservation techniques.Although there was an average increase in dressing ability (measured by the NottinghamStroke Dressing Assessment³) of 11% in the treatment group, it is possible that an optimal improvementin dressing ability was not achieved as there was no systematic approach to the assessment,analysis of the underlying problems or targeted treatment of cognitive difficulties.Similarly, although two other studies have reported improvements in dressing performancefollowing task-specific interventions,^8,9 the underlying cognitive impairments associated with dressing remainunexplored.

A subsequent single-blind randomized multiple-baseline experiment¹⁰ combined naturalistic observation ofdressing abilities, systematic neuropsychological assessment and administration of targeteddressing interventions. Results demonstrated that there was a significant treatment effect,measured by the Nottingham Stroke Dressing Assessment and observation-based t-shirt test,for those inpatients with right hemisphere stroke. There was, however, no therapy-relatedimprovement for those with left or bilateral damage and apraxia. Because of the small numberof case studies in this experiment (n = 8), a further study was required toestablish the potential benefits of this approach over conventional approaches employed byoccupational therapists. In following the MRC framework for the development and evaluationof complex interventions¹¹ a pilot phase II randomized controlled trial was designed. Our aim was toconduct an evaluation of two approaches used to treat stroke patients with persistentdressing problems and accompanying cognitive difficulties after stroke.

Methods

Consecutive inpatients on the stroke rehabilitation wards at Nottingham UniversityHospitals Trust were monitored by the ward occupational therapists to identify thosepatients with persistent dressing difficulties. Patients were deemed suitable for the studyif they had received two weeks of conventional rehabilitation and still required help todress. Patients were invited to take part in the study if they were impaired (scored lessthan 100% maximum score) on the Nottingham Stroke Dressing Assessment^3,12 and on one or more items in a briefcognitive screening test: line cancellation¹³ to detect visual neglect (maximum score36, impaired <33); 10-hole peg transfer test¹⁴ with the non-paretic hand to detectdexterity problems not due to paresis (impaired >22 seconds); the Object Decision subtestfrom the Visual Object and Space Perception assessment¹⁵ (maximum score 20, impaired <12);Gesture Imitation to detect apraxia¹⁶ (maximum score 20, impaired <15). To ensure selection of patientsable to participate in dressing practice, the exclusion criteria included the inability totolerate sitting in a chair for 15 minutes, premorbid disability (Rankin¹⁷ >3) and known diagnosisof depression or dementia.

In terms of comprehension, patients had to be able to understand English if it wasn’t theirfirst language. The Sheffield Aphasia Test¹⁸ was used to assess aphasia and adaptedinformation and consent forms were used where appropriate. Demographic data were alsocollected on the Barthel Index,¹⁹ Motricity Index,²⁰ age, gender and side of stroke.

Following baseline assessments and using concealed allocation via the University ofNottingham Clinical Trials Unit internet randomization service, patients were randomized toone of two treatment groups; conventional occupational therapy (the ‘Functional approach’)or the ‘Neuropsychological approach.’ Patients were stratified by side of stroke andseverity of their dressing problem as measured by the Nottingham Stroke Dressing Assessmentscore. The two groups continued with their usual rehabilitation therapy and nursing care andonly differed in the type of dressing practice provided by the trial occupationaltherapists. Both interventions were delivered by two research occupational therapistsexperienced in the treatment of stroke patients.

As side of stroke was initially recorded from the medical notes, brain scans were laterreviewed by an experienced stroke radiologist. This identified a subgroup of patients withdefinite or probable bilateral hemisphere damage and these were treated as a separatecategory in analysis of results.

Interventions were prescribed according to group allocation. Treatment manuals hadpreviously been developed for both dressing approaches using comprehensive literaturesearches, survey results⁶ and occupational therapy text books.²¹ Patients allocated to the functionalapproach were given repeated dressing practice using a problem-solving approach, withassistance when required. This approach is commonly used by occupational therapists in theUK and has been shown to have a beneficial effect on dressing performance.⁷ Dressing interventions wouldinclude components such as putting the affected arm into the sleeve first, crossing affectedleg over other leg to reach feet, energy conservation techniques, etc. There was no attemptto formally assess the patient’s cognitive difficulties or relate them to evidence on whichapproach to training might be the most successful.

Patients assigned to the neuropsychological approach received further detailed cognitivetesting and an assessment of the impact of cognitive deficits on dressing by observation ofa standard task of putting on a t-shirt,¹⁰ with performance scored using an erroranalysis rating form.¹² This error analysis identified the presence of cognitive problems such asimpaired attention, spatial confusion and action sequence errors. On the basis of the testresults and the types of error observed, treatment interventions were selected from a menuof evidence-based techniques described in the pre-prepared neuropsychological treatmentmanual. The most commonly used specific techniques were cueing and alerting procedures tocombat neglect or attentional difficulties,^22–24 systematic laying out of clothing to reduce spatialconfusion,¹⁰ andgraded errorless learning strategies to enhance acquisition of dressing skills.²⁵ Fidelity of treatment inboth patient groups was monitored by an independent researcher who observed random dressingsessions to ensure the manuals were adhered to.

The optimal intensity of dressing practice as indicated in our previous single caseexperiments¹⁰dictated that patients in both groups received dressing treatment three times a week for aperiod of six weeks. Patients continued to receive dressing treatment in their own home ifthey were discharged from hospital before the end of the treatment period.

Patients were assessed at six weeks after randomization by an independent assessor who wasmasked to the patient’s treatment group allocation. Masking of the independent assessor wasmonitored by completion of a best guess form. All patients were assessed on the NottinghamStroke Dressing Assessment^3,12 and thecognitive tests which had been used in initial screening were repeated (linecancellation¹³ todetect visual neglect; 10-hole peg transfer test¹⁴ with the non-paretic hand to detectdexterity problems not due to paresis; the Object Decision subtest from the Visual Objectand Space Perception assessment;¹⁵ Gesture Imitation to detect apraxia¹⁶). Performance on the Nottingham StrokeDressing Assessment was selected as the primary outcome measure for the study.

As this was a feasibility study, no formal power calculation was performed. However basedon a similar dressing study⁷ using the same primary outcome measure, it was estimated that the studywould require 35 patients per group (80% power to detect an effect at the 0.05 level).Although the previous trial had investigated a slightly different approach to dressingdifficulties, the authors felt that a pragmatic study of this magnitude would be informativefor the proposed trial. Statistical analyses included Student’s t-tests ofwithin group means and standard deviations over time, between group differences at six weeksand planned subgroup analyses for patients with right hemisphere and those with left orbilateral damage were also carried out.

Results

Of the 965 patients screened during the trial (1 March 2008–28 February 2010), we soughtconsent from 110. Of these, 40 passed the screening tests. The remaining 70 patients (64%)were randomized to either the neuropsychological group (n = 36) or to thefunctional group (n = 34). A sample of treatment sessions were monitored toensure that they included the actual treatment prescribed in the manual. We found a highlevel of fidelity of treatment in both treatment groups. Masking of the outcome assessor wastested and found to be compromised for only six patients.

Figure 1 shows patient selectionand drop-outs.

Table 1 shows the details ofthose who completed the trial and Table2 shows their scores on the baseline assessments. The treatment groups were wellmatched on all variables.

The interventions provided in both arms of the study were well tolerated and found to beacceptable to patients. (Due to space limitations these findings on acceptability will bereported in detail in a further paper.) The number of treatment sessions delivered to eachgroup during the six-week period was well matched. The Neuropsychological group received amedian of 13 sessions (min 0, max 18) and the Functional group received a median number of12 sessions (min 0, max 18). The key reasons for people not receiving all 18 sessions were:the patient deteriorated or died, moved out of the geographical catchment area, nursingstaff had dressed the patient before the therapist arrived or the patient had reachedindependent dressing before the end of the six-week intervention period.

Performance at the outcome assessments is shown in Table 3. Compared to the baseline assessments, bothtreatment groups showed significant improvements in dressing ability (improvements of 31%and 22% on the Nottingham Stroke Dressing Assessment for the Neuropsychological andFunctional groups respectively) but the groups did not differ significantly in this respect(t(62) = 1.3, NS).

Table 3 also demonstrates thatthe cognitive tests showed a trend towards improvement from baseline. The Neuropsychologicalgroup showed a greater reduction in visual neglect on the cancellation test than theFunctional group (t(62) = 2.1, P < 0.05). There were nosignificant group differences on any other cognitive test.

Following the approach used in our previous single-case design investigation,⁸Table 4 and Figure 2 show planned subgroup analyses for patientswith right hemisphere and those with left or bilateral damage. These show a trend towardsgreater dressing improvement with the neuropsychological approach for those with righthemisphere damage (P = 0.07, one-tailed) who also show a significantlygreater reduction of visual neglect on the line cancellation test. In contrast, the subgroupwith left or bilateral damage show no differences between treatment approaches close tosignificance, either in dressing or test performance.

Discussion

Our feasibility randomized controlled trial indicates that both groups improved in theirdressing performance over the trial intervention period. We found no statisticallysignificant differences between the groups on performance on our main outcome measure: theNottingham Stroke Dressing Assessment. However we did find a trend for improvement in thosepatients with right hemisphere stroke who were receiving the systematic neuropsychologicalapproach to dressing. This outcome supports our previous findings in this patientgroup¹⁰ and addsweight to the hypothesis that patients with right hemisphere stroke with accompanyingdressing problems and cognitive deficits can be treated successfully using a systematicneuropsychological approach. The fact that these patients also demonstrated a significantlygreater reduction in visual neglect suggest that it was the use of techniques to reduceneglect which had the greatest impact on dressing ability in this group.

It is likely that a stronger treatment effect would be observed in a purpose designed trialaimed specifically at patients with right hemisphere damage. The current study included allpatients with cognitive impairment and dressing difficulties, with no preselection for siteof brain lesion. This meant that the subgroup of right hemisphere patients allocated to theneuropsychological was small. Furthermore, the techniques described in our treatment manualwere culled from the wider neuropsychological literature and seldom specifically describedthe application of these treatments to patients with dressing difficulties.^22–25 Our experience ofthe trial will allow us to write a more targeted treatment manual, which together with alarger sample size may demonstrate a more substantial treatment effect. In addition, weincluded two intervention groups in our dressing trial and did not include a conventionalcontrol group. Our reasoning behind the inclusion of a manualized functional approach wasbased on the positive benefits found in a similar interventional dressing study.⁷ Although the approach weapplied was based on routine care provided by occupational therapists in the UK, the contentof the functional manual was very prescriptive and the frequency with which the interventionwas delivered was much greater than that routinely provided on the stroke rehabilitationwards. We therefore believe that a clinically significant difference may have been observedon the main outcome measure had we included a conventional control group.

The left and bilateral damage subgroup showed no sign of any benefit of theneuropsychological approach. One reason for this may be that their dressing difficultieswere simply less severe and therefore that there was less room for improvement. For righthemisphere-damaged patients, the presence of unilateral neglect, spatial confusion or poorsustained attention have a devastating effect on dressing skills.³ In contrast, left hemisphere damage isassociated with apraxia and related problems in control of action, but the impact of theseon everyday functioning is more subtle and seems to depend on the exact cognitive demands ofany given task situation.²⁶ This said, there were left hemisphere damaged apraxic patients in thisstudy who were unable to regain independence in dressing, and reasons for this lack ofbenefit of the neuropsychological approach need to be considered. A likely explanation isthat the techniques to assist them are not well developed: there are only few, small-scaleexperimental studies on intervention for apraxia.²⁷ This contrasts with the well-developedliterature on intervention for visual neglect.²⁸ Our results suggest that evidence-basedsystematic interventions tailored to the deficits found in right hemisphere-damaged strokepatients are likely to be beneficial.

The main weakness of our study was its relatively small sample size which limited the powerto detect a statistically significant effect on dressing performance. Similarly, althoughthe subgroup analysis was planned, it was nonetheless carried out on a small number ofpatients. However, we believe that the sample size achieved has allowed us to demonstratethat this approach to dressing is a feasible method to employ with stroke patientsexperiencing persistent dressing difficulties and such an intervention can be carried out ona busy stroke rehabilitation unit. We believe the indication of possible benefit in righthemisphere stroke is worthy of further enquiry and should be tested in a multicentre trialutilizing the findings from this study.

Acknowledgments

We would like to thank the patients who participated in the study, the staff on the strokerehabilitation wards at Nottingham University Hospitals Trust.

Footnotes

References