Upper Limb Outcome Measures Used in Stroke Rehabilitation Studies: A Systematic Literature Review

Abstract

Background: Establishing which upper limb outcome measures are most commonly used in stroke studies may help in improving consensus among scientists and clinicians.

Objective: In this study we aimed to identify the most commonly used upper limb outcome measures in intervention studies after stroke and to describe domains covered according to ICF, how measures are combined, and how their use varies geographically and over time.

Methods: Pubmed, CinHAL, and PeDRO databases were searched for upper limb intervention studies in stroke according to PRISMA guidelines and477 studies were included.

Results: In studies 48different outcome measures were found. Only 15 of these outcome measures were used in more than 5% of the studies. The Fugl-Meyer Test (FMT)was the most commonly used measure (in 36% of studies). Commonly used measures covered ICF domains of body function and activity to varying extents. Most studies (72%) combined multiple outcome measures: the FMT was often combined with the Motor Activity Log (MAL), the Wolf Motor Function Test and the Action Research Arm Test, but infrequently combined with the Motor Assessment Scale or the Nine Hole Peg Test. Key components of manual dexterity such as selective finger movements were rarely measured. Frequency of use increased over a twelve-year period for the FMT and for assessments of kinematics, whereas other measures, such as the MAL and the Jebsen Taylor Hand Test showed decreased use over time. Use varied largely between countries showing low international consensus.

Conclusions: The results showed a large diversity of outcome measures used across studies. However, a growing number of studies used the FMT, a neurological test with good psychometric properties. For thorough assessment the FMT needs to be combined with functional measures. These findings illustrate the need for strategies to build international consensus on appropriate outcome measures for upper limb function after stroke.

Fig 1. Flow-chart illustrating search strategy and number of studies selected.

Fig 2. Frequency of use of different upper limb outcome measures (in % of studies).

Frequency of use varies widely, between 36% and 1%. Only 15 measures were used in more than 5% of studies (dotted line). The 48 outcome measures are in alphabetic order: AMAT = Arm Motor Ability Test, ARAT = Action Research Arm Test, Ashworth = Ashworth scale, BBT = Box and Blocks Test, CAHAI = Chedoke Arm Hand Inventory, CMSA = Chedoke McMaster Stroke Assessment, COPM = Canadian Occupational Performance Measure, DAS = Disability Assessment Scale, DTI = Diffusion Tensor Imaging, EMG = Electromyography, FAT = Frenchay Arm Test, FC = Force Control, fMRI = Functional Magnetic Resonance Imaging, FMT = Fugl-Meyer Test, FTHUE = Functional Test for the Hemiplegic Upper Extremity, FTT = Finger Tapping Test, GOT = Grating Orientation Task, GRT = Grasp Release Test, HFS = Hand Function Survey, HFT = Hand Function Test, JTHT = Jebsen Taylor Hand Test, KIN = Kinematics, MAL = Motor Activity Log, MAM36 = Manual Ability Measurement 36, MAS = Motor Assessment Scale, MHS = Mini Hand Score, MI = Motricity Index, MMDT = Minnesota Manual Dexterity Test, NHPT = Nine Hole Peg Test, NSA = Nottingham Sensory Assessment, PT = Pegboard Test, RELHFT = Rehabilitation Engineering Laboratory Hand Function Test, RMA = Rivermead Motor Assessment, ROM = Range of Movement, SHFT = Shollerman Hand Function Test, SHPT = Sixteen Hole Peg Test, SIAS = Stroke Impairment Assessment Set, SMES = Sodring Motor Evaluation Scale, SSDI = Standardized Somatosensory Deficit Index, STEF = Simple Test for Hand Function, TDT = Tactile Discrimination Test, TMS = Transcranial Magnetic Stimulation, TS = Tardieu Scale, UEFT = Upper Extremity Function Test, ULIS = Upper Limb Impairment Scale, VAS = Visual Analogue Scale, VFHT = Von-Frey Hair Test, WMFT = Wolf Motor Function Test.

Fig 3. Frequency of use (%) of outcome measures according to ICF domains (A, B) and advanced methods (C).

Abbreviations as in Fig 2.

Fig 4. Combination of outcome measures.

Measures related to the ICF Activity level are listed on the horizontal, those related to ICF Body function are shown on the left-top side, and those qualified as ‘Advanced methods’ are shown on the right-top side of the triangle. A curved link (line) between two different measures indicates their combined use within a study. The thickness of the curved line represents the frequency of occurrence across studies of a given combination.

Fig 5. Increasing and decreasing frequency of use of outcome measures.

The FMT and KIN both showed significantly increasing trends of use according to Mann-Kendall test (P<0.05, two-tailed). The MAL and the JTHT showed decreasing trends of use (P<0.05, two-tailed).

Fig 6. Frequency of use in the ten countries with most publications.

Distributions (in % use) shown for Ashworth, Motor Activity Log (MAL), Fugl-Meyer Test (FMT) and Action Research Arm Test (ARAT). Frequency of use of particular measures across countries varied greatly and was not homogenous. The Fugl-Meyer Test (FMT, Fig 5C) has high rates of use in most countries, except in Australia and UK, where the Action Research Arm Test (ARAT, Fig 5D) was used more often.