Research Article | Open Access
Pre- and Postintervention Factor Structure of Functional Independence Measure in Patients with Spinal Cord Injury
Objective. To evaluate the factor structure of Functional Independence Measure (FIM®) scale amongst people with spinal cord injury (SCI). Methods. This was a retrospective, register-based cohort study on 155 rehabilitants with SCI. FIM was assessed at the beginning and at the end of multidisciplinary inpatient rehabilitation. The internal consistency of the FIM was assessed with Cronbach’s alpha and exploratory factor analysis was employed to approximate the construct structure of FIM. Results. The internal consistency demonstrated high Cronbach’s alpha of 0.95 to 0.96. For both pre- and postintervention assessments, the exploratory factor analysis resulted in 3-factor structures. Except for two items (“walking or using a wheelchair” and “expression”), the structures of the identified three factors remained the same from the beginning to the end of rehabilitation. The loadings of all items were sufficient, exceeding 0.3. Both pre- and postintervention chi-square tests showed significant values 0.0001. The “motor” domain was divided into two factors with this 2-factor structure enduring through the intervention period. Conclusions. Amongst rehabilitants with SCI, FIM failed to demonstrate unidimensionality. Instead, it showed a 3-factor structure that fluctuated only little depending on the timing of measurement. Additionally, when measured separately, also motor score was 2-dimensional, not 1-dimensional. Using a total or subscale FIM, scores seem to be unjustified in the studied population.
The Functional Independence Measure (FIM) was developed in the 1980s–1990s as a scale measuring a need for assistance in the medical system in the United States . Since then, FIM has been extensively used around the world for diverse health disorders [2–10].
The best known FIM outcome is a radar diagram showing the distribution of individual scores obtained from 18 items of FIM . In addition to this graphical output, total scores and scores for two subscales (“motor” an “cognitive”) are used to describe the severity of disability and changes in functioning over time. The usefulness of these scores has been questioned [5, 12–17]. To be reliable, a score should represent a unified outcome on an interval scale. This means that a scale must represent only one underlying construct, that is, pain or disability severity, to be able to produce a reliable numeric score of several items. This unidimensionality of FIM has not been established yet. Instead, numerous studies have reported that FIM is a multidimensional scale and, because of this multidimensionality, aggregated numeric scores may be meaningless [5, 12–17]. Such studies have been conducted amongst stroke survivors, elderly patients, and SCI patients amongst others. While the multidimensionality of FIM has become well known, there is no knowledge if its factor structure prevails the same over time. To make a comparison between reliable repeated measurements, a scale must preserve its factor structure over time. In other words, the factor structures of FIM should be similar regardless of the timing of measurement, at the beginning or at the end of intervention or even years after. This kind of structure stability of FIM has not been studied so far.
Even if the FIM has been widely used amongst people with SCI [3, 4, 8, 9, 18–25], previous evidence on FIM validity in SCI is scarce. A few studies have reported that FIM may have a substantial ceiling effect in the cognitive subscale and a possible item redundancy within the motor subscale [20, 21] with poor psychometric properties assessed with the Rasch analysis . Based on the Rasch analysis, the FIM motor score was not recommended for use, not at least as a raw sum of item scores. The cognitive scale was found to work well except for the substantial ceiling effect . Due to the uncertainty concerning the reliability of FIM amongst people with SCI, other scales have been suggested to replace FIM in that population, such as Spinal Cord Independence Measure (SCIM) .
The objective of this study was to evaluate the factor structure of FIM amongst SCI rehabilitants.
This was a retrospective, register-based cohort study amongst 155 patients with SCI who participated in a multidisciplinary inpatient rehabilitation in a university hospital clinic between January 2010 and April 2017. The FIM was filled out by nurses certified to use FIM except for walking at stairs, which was assessed by a trained physiotherapist. The first assessment was usually performed within two days after the admission. In case of a register-based study, no approval of an ethical committee was needed.
In Finland, the rehabilitation of stroke is arranged in several different standardized phases. Subacute rehabilitation is separated from the acute treatment and rehabilitation of SCI. Thus, in this study on subacute rehabilitation, the patients were admitted for multidisciplinary rehabilitation usually after a few months after onset. The main goal of rehabilitation team (a physician, a nurse, a physiotherapist, an occupational therapist, a psychologist, and a social worker) at that subacute stage is to map the situation and prepare the return of a patient to his or her community—even if it is going to happen much later after next several months. After relatively short (a few weeks) stay at a hospital rehabilitation ward, patients with SCI are transferred to one of the five SCI centers for a longer rehabilitation that usually takes a few months. The rehabilitation ward also arranges rehabilitation for patients with SCI who already have returned to their homes.
The FIM has been broadly described previously . The FIM consists of 18 items. Each item is valued on a Likert-like scale from one (“total assistance in all areas”) to seven (“completely independent”). Three sum scores are calculated and documented in a patient’s record: a “motor” score (the sum of the scores collected from the first 13 items), a “cognitive” score (the sum of the scores collected from the last five items), and a total score, that is, the sum of motor and cognitive scores. The maximum total score is 126 points, while the minimum is 18 points. The FIM items can be seen in Tables 1 and 2.
2.1. Statistical Analysis
The internal consistency of the FIM was assessed with Cronbach’s alpha along with its one-sided 95% confidence limit (95% CI). Alpha ≥ 0.9 was considered excellent, 0.8 good, 0.7 acceptable, 0.6 questionable, 0.5 poor, and 0.5 was considered unacceptable. This study employed exploratory factor analysis to approximate the construct structure of FIM. The goal was to determine if FIM measures only one latent trait (e.g., disability or need for assistance) or if there are other possible significant latent variables affecting the results as well. The results were analyzed both numerically and graphically. Exploratory factor analysis (principal factors) was applied with a minimum eigenvalue for retention set at >1.0 (Kaiser’s rule). The results of factor analysis were rotated by an orthogonal Varimax rotation, assuming that there is no correlation between factors. Retained and excluded factors were also explored visually on a scree plot including a parallel analysis. The demographic characteristics were reported as percentages when appropriate. Otherwise, means, ranges, standard deviations (SDs), and interquartile ranges (IQRs) were reported. All the analyses were performed using Stata/IC Statistical Software, Release 14, StataCorp LP (College Station, TX, USA).
The median age of the 155 patients was 61 (IQR: 50 to 69) years (Table 3). Of them, 87 (56%) were men and 68 (44%) were women (Table 2). Of these patients, 63 (41%) had paraplegia, 67 had tetraplegia (43%), and 25 (16%) could not be exactly specified by the retrospective data. The multidisciplinary rehabilitation program started (median) 3.1 (IQR: 1.2 to 13.1) months after the SCI onset. The median duration of rehabilitation was 17 (IQR: 11 to 32) days. Therapy and training were started at the day of admission. FIM measurements were conducted (median) within 2 days (IQR: 1 to 3) after admission and within 1 day (IQR: 0 to 2) before the day of discharge from the rehabilitation ward. The median FIM total scores were 107 points at the beginning and 114 points at the end of rehabilitation.
|aAll of the estimates were distributed abnormally.|
The internal consistency demonstrated high Cronbach’s alpha of 0.95 (95% CI: 0.94) before and 0.96 (95% CI: 0.95) after the intervention.
For both pre- and postintervention assessments, the exploratory factor analysis resulted in 3-factor structures (Figure 1 and Table 1). The eigenvalues for these three factors were 10.0, 1.7, and 1.2 when assessed at the beginning and 10.5, 1.5, and 0.9 at the end of rehabilitation. Even if the third factor had a posttest eigenvalue below Kaiser’s cutoff of 1.0, the parallel analysis of scree plot was considered to produce a more precise and logically acceptable result.
Except for two items (“walking or using a wheelchair” and “expression”), the structures of the identified three factors remained the same at the beginning and at the end of rehabilitation. The item “walking or using a wheelchair” migrated from factor 2 to factor 1, and the item “expression” from factor 3 to factor 2. The loadings of all the items were sufficient, exceeding 0.3 . Both pre- and postintervention chi-square tests showed significant values 0.0001.
The separate factor analysis of only motor score revealed a 2-factor structure in both pre- and posttesting. These two factors had eigenvalues of 8.96 and 1.36 in pretest analysis and 9.13 and 1.22 in posttest analysis, respectively. First factor with much larger eigenvalue was connected to bathing, dressing lower body, toileting, sphincter control, transfers, and using stairs. In turn, second factor was related to eating, grooming, dressing upper body, and moving by walking or moving using a wheelchair. These associations prevailed from pre- to posttesting.
In this retrospective register-based study amongst SCI patients who participated in multidisciplinary inpatient rehabilitation, four main results were observed. First, FIM scale was not unidimensional. Instead, it demonstrated a 3-factor structure. Second, this 3-factor structure was observed by both repeated measurements, at the beginning and at the end of rehabilitation. Third, there were small fluctuations in the intrinsic constructions of these factors: the loadings of some items varied at different time points. Fourth, while so-called “cognitive” subscale of FIM formed (as could be expected) one relatively stable factor, “motor” subscale unexpectedly disintegrated into two distinct factors at both pre- and postintervention time points.
Even if the results were statistically significant, the sample of 155 patients may be too small to examine an 18-item scale and to represent all people with SCI in general. The time interval of two or three weeks between the repeated measures was probably too short to detect small changes in FIM scores or the FIM factor structure. This limitation might also affect the psychometric properties of FIM as the assessors, when scoring the final situation, almost certainly still recalled the initial scores resulting in reduced test-retest reliability.
The multidimensional nature of FIM observed in this study was not a surprise, as previous studies have expressed their concern about the multidimensionality of FIM and, as a result, the uncertainty in calculating the FIM total score [5, 12–17]. Indeed, the total score that measures more than one construct is not as important as was originally thought. We, however, expected at least to see the 2-dimensionality familiar to all FIM users: “motor” and “cognitive” domains. These two dimensions were not confirmed. Instead, three different factors were observed. This kind of multidimensionality has been reported before in elderly people with multiple causes of disability , but this phenomenon has not been described earlier amongst people with SCI.
Except for the item “expression” (which migrated between two factors during the intervention), the cognitive domain remained relatively stable including, as expected, last five items of FIM. Instead, the “motor” domain was divided into two factors. The motor domain’s 2-factor structure endured through the intervention period with small fluctuations. When reporting problems with factor structure of FIM total score, previous studies usually assumed that, when evaluating separately, motor and cognitive scales should demonstrate unidimensionality. That was not the case in the present study when motor scale was apportioned into two factors. This finding may signal that caution should be applied not only when dealing with the FIM total score but also when interpreting its motor score separately.
The results of the present study yield unambiguous and clinically significant suggestions. As a unidimensional scale, the cognitive subscale of FIM seemed to be able to measure reliably the need for assistance in cognitive functioning. However, the cognitive subscale of FIM could be used in SCI patients with some reservations, as it has to be taken into account that its intrinsic factor structure may fluctuate depending on the timing of measurement. The results are questioning the use of a FIM total or motor score within the studied population as both constructs were found to be multidimensional and, therefore, they may be unable to demonstrate an outcome measured on a unified interval scale.
These considerations do not mean that FIM should not be used at all. FIM has been widely used for 30 years, and the scale holds the status of a worldwide standard assessment of disability (or to be more precise, the assessment of a need for assistance). Such scales are rare and are of a great value as they produce the results comparable across different settings, disorders, and cultures. Our results suggest, however, that (in the studied population of rehabilitants with SCI) a total score or subscores have only limited values. Instead, we would recommend using individual scores obtained from each of 18 items of FIM assessment and report these results as a radial diagram familiar to all FIM users. We also suggest employing informal descriptions in patients’ reports to define the restrictions more widely.
Further research conducted on larger samples in different settings and with a longer time interval between repeated measurements may amplify our findings. Considering the widespread use of FIM, the need for such research is apparent. The validity of FIM scores is important not only to health professionals but also to stakeholders when weighing the effectiveness and efficiency of a particular intervention based on the improvements measured by FIM scores.
Amongst rehabilitants with SCI, FIM failed to demonstrate unidimensionality. Instead, it showed a 3-factor structure that fluctuated only little depending on the timing of measurement. Additionally, when measured separately, also motor score was 2-dimensional, not 1-dimensional. Using a total or subscale FIM, scores seem to be unjustified in the studied population.
(i) The study used data gathered during a long time interval of seven years. (ii) The comprehensive analysis of the FIM psychometric properties inferred probable instability of the FIM factor structure over time, a topic that has been studied only little so far. (iii) Even if the results were statistically significant, the sample of 155 patients may be too small to examine an 18-item scale and to represent all people with SCI in general. (iv) The time interval of two or three weeks between the repeated measures was probably too short to detect small changes in FIM scores or the FIM factor structure. (v) This limitation might also affect the psychometric properties of FIM as the assessors, when scoring the final situation, almost certainly still recalled the initial scores resulting in reduced test-retest reliability.
The study protocol and statistical analysis plan are available on request from the corresponding author.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Mikhail Saltychev, Janne Lähdesmäki, Petteri Jokinen, and Katri Laimi substantially contributed to conception and design, acquisition, and interpretation of data, revised the work, and finally approved the version to be published. Mikhail Saltychev took responsibility for the statistical analysis of data.
- A. Torres, “Capturing Functional Independence Measure (FIM,” Rehabilitation Nursing, 2016.
- J.-J. Dubost, I. Fis, P. Denis et al., “Polyarticular septic arthritis,” Medicine, vol. 72, no. 5, pp. 296–310, 1993.
- R. J. Marino, M. Huang, P. Knight, G. J. Herbison, J. F. Ditunno, and M. Segal, “Assessing selfcare status in quadriplegia: Comparison of the quadriplegia index of function (QIF) and the functional independence measure (FIM),” Paraplegia, vol. 31, no. 4, pp. 225–233, 1993.
- M. E. Segal, J. F. Ditunno, and W. Staas, “Interinstitutional agreement of individual functional independence measure (Fim) items measured at two sites on one sample of sci patients,” Paraplegia, vol. 31, no. 10, pp. 622–631, 1993.
- S. J. Cavanagh, K. Hogan, V. Gordon, and J. Fairfax, “Stroke-specific FIM models in an urban population,” The Journal of Neuroscience Nursing : Journal of the American Association of Neuroscience Nurses, vol. 32, no. 1, pp. 17–21, 2000.
- I. Hong, H.-S. Woo, S. Shim, C.-Y. Li, L. Yoonjeong, and C. A. Velozo, “Equating activities of daily living outcome measures: the Functional Independence Measure and the Korean version of Modified Barthel Index,” Disability and Rehabilitation, pp. 1–8, 2016.
- M. Invernizzi, S. Carda, P. Milani et al., “Development and validation of the Italian version of the Spinal Cord Independence Measure III,” Disability and Rehabilitation, vol. 32, no. 14, pp. 1194–1203, 2010.
- G. Lawton, Å. Lundgren-Nilsson, F. Biering-Sørensen et al., “Cross-cultural validity of FIM in spinal cord injury,” Spinal Cord, vol. 44, no. 12, pp. 746–752, 2006.
- L. H. Lugo, F. Salinas, and H. I. García, “Out-patient rehabilitation programme for spinal cord injured patients: Evaluation of the results on motor FIM score,” Disability and Rehabilitation, vol. 29, no. 11-12, pp. 873–881, 2007.
- S. Naghdi, N. N. Ansari, P. Raji, A. Shamili, M. Amini, and S. Hasson, “Cross-cultural validation of the Persian version of the Functional Independence Measure for patients with stroke,” Disability and Rehabilitation, vol. 38, no. 3, pp. 289–298, 2016.
- The FIM® instrument: its background, structure, and usefulness, Uniform Data System for Medical Rehabilitation, Buffalo, NY, USA, 2012, http://www.udsmr.org/Documents/The_FIM_Instrument_Background_Structure_and_Usefulness.pdf.
- H. G. Dickson and F. Kohler, “The multi-dimensionality of the FIM motor items precludes an interval scaling using rasch analysis,” Scandinavian Journal of Rehabilitation Medicine, vol. 28, no. 3, pp. 159–162, 1996.
- P. Gerrard, R. Goldstein, M. A. Divita et al., “Validity and reliability of the FIM instrument in the inpatient burn rehabilitation population,” Archives of Physical Medicine and Rehabilitation, vol. 94, no. 8, pp. 1521–1526, 2013.
- N. Pollak, W. Rheault, and J. L. Stoecker, “Reliability and validity of the FIM for persons aged 80 years and above from a multilevel continuing care retirement community,” Archives of Physical Medicine and Rehabilitation, vol. 77, no. 10, pp. 1056–1061, 1996.
- J.-F. Ravaud, M. Delcey, and A. Yelnik, “Construct validity of the functional independence measure (FIM): questioning the unidimensionality of the scale and the ‘value’ of FIM scores,” Journal of rehabilitation medicine, vol. 31, no. 1, pp. 31–41, 1999.
- B. D. Wright, J. M. Linacre, R. M. Smith, A. W. Heinemann, and C. V. Granger, “FIM measurement properties and Rasch model details,” Scandinavian Journal of Rehabilitation Medicine, vol. 29, no. 4, pp. 267–272, 1997.
- E. Yasąr, S. Kesikburun, and B. Yilmaz, “Functional assessment of SCI patients by FIM: Yes or no,” Spinal Cord, vol. 53, no. 12, p. 894, 2015.
- M. Beninato, K. S. O'Kane, and P. E. Sullivan, “Relationship between motor FIM and muscle strength in lower cervical-level spinal cord injuries,” Spinal Cord, vol. 42, no. 9, pp. 533–540, 2004.
- R. K. Bode, A. W. Heinemann, A. J. Kozlowski, and C. R. Pretz, “Self-scoring templates for motor and cognitive subscales of the FIM instrument for persons with spinal cord injury,” Archives of Physical Medicine and Rehabilitation, vol. 95, no. 4, pp. 676–e5, 2014.
- K. M. Hall, M. E. Cohen, J. Wright, M. Call, and P. Werner, “Characteristics of the functional independence measure in traumatic spinal cord injury,” Archives of Physical Medicine and Rehabilitation, vol. 80, no. 11, pp. 1471–1476, 1999.
- A. I. Masedo, M. Hanley, M. P. Jensen, D. Ehde, and D. D. Cardenas, “Reliability and validity of a self-report FIM™ (FIM-SR) in persons with amputation or spinal cord injury and chronic pain,” American Journal of Physical Medicine & Rehabilitation, vol. 84, no. 3, pp. 167–176, 2005.
- R. McConnell, J. Y. Zheng, E. H. Chadd, and C. Kalpakjian, “Poster 79 comprehension of visual displays of the functional independence measure (Fim) in patients with spinal cord injury in acute rehabilitation,” PM & R : The Journal of Injury, Function, and Rehabilitation, vol. 8, Supplement 187, no. 9, 2016.
- J. W. Middleton, L. A. Harvey, J. Batty, I. Cameron, R. Quirk, and J. Winstanley, “Five additional mobility and locomotor items to improve responsiveness of the FIM in wheelchair-dependent individuals with spinal cord injury,” Spinal Cord, vol. 44, no. 8, pp. 495–504, 2006.
- C. R. Pretz, A. J. Kozlowski, S. Charlifue, Y. Chen, and A. W. Heinemann, “Using rasch motor FIM individual growth curves to inform clinical decisions for persons with paraplegia,” Spinal Cord, vol. 52, no. 9, pp. 671–676, 2014.
- S. Warschausky, J. B. Kay, and D. G. Kewman, “Hierarchical linear modeling of FIM instrument growth curve characteristics after spinal cord injury,” Archives of Physical Medicine and Rehabilitation, vol. 82, no. 3, pp. 329–334, 2001.
- L. A. Harvey and K. D. Anderson, “The Spinal Cord Independence Measure,” Journal of Physiotherapy, vol. 61, no. 2, p. 99, 2015.
- A. B. Costello and J. W. Osborne, “Best practices in exploratory factor analysis: four recommendations for getting the most from your analysis,” Practical Assessment, Research and Evaluation, vol. 10, no. 7, pp. 1–9, 2005.
- C. Glenny and P. Stolee, “Comparing the Functional Independence Measure and the interRAI/MDS for use in the functional assessment of older adults: A review of the literature,” BMC Geriatrics, vol. 9, no. 1, article no. 52, 2009.
Copyright © 2017 Mikhail Saltychev et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.