The reliability of the modified lower extremity functional scale among adults living with HIV on antiretroviral therapy, in Rwanda, Africa

Abstract Peripheral neuropathy (PN) is common among people living with HIV (PLHIV) on antiretroviral therapy (ART), and affects their daily functional ability and quality of life. Lower extremity functional ability, which is most commonly compromised in patients with PN, has not been clearly evaluated in an African setting, with regard to functional limitations. The lower extremity functional scale (LEFS) was originally developed and validated among elderly people in the USA, where the environment and activities of daily life are very different from those in Rwanda. The purpose of this study was to adapt and establish the reliability of LEFS, among adults living with HIV on ART, in a Rwandan environment. The study translated LEFS from English to Kinyarwanda, the local language spoken in Rwanda, the LEFS was then modified accordingly, and tested for test-retest reliability among 50 adult PLHIV on ART. An average Spearman rank order correlation coefficient, ρ ≥ 0.7, was considered optimal for reliability. Prior to the modification of the LEFS and in the initial testing of the translated LEFS, none of the activities was strongly correlated (ρ ≥ 0.8); most of the activities (90%, 18/20) were moderately correlated (ρ ≥ 0.5) and 10% (2/20) were weakly correlated (ρ ≤ 0.5). The ρ of most of the functional activities improved after modification by an expert group to ρ ≥ 0.7, establishing reliability and validity of LEFS among PLHIV on ART with lower extremity functional limitations, in this environment. In conclusion, this study demonstrated the importance of modifying and establishing test – retest reliability of tools derived from developed world contexts to local conditions in developing countries, such as in Rwanda. The modified LEFS in this study can be used in Rwanda by clinicians, specifically at ART clinics to screen and identify people with functional limitations at an early stage of the limitations, for treatment, rehabilitation and/or referral to appropriate health care services.

The LEFS, a tool for evaluating lower limb functional ability, was developed and validated (Binkley, Stratford, Lott & Riddle 1999) in the USA where the environment and activities of daily life are different from those in developing countries such as Rwanda. The activities in developed countries are more urbanised, for instance structured sporting activities, while in developing countries they are more rural, including farming and agricultural activities. To our knowledge, this tool has not been validated in any African country. The purpose of this study was to re-establish the reliability of the modified LEFS so as to assess functional ability of the lower extremity among adults living with HIV and on ART in Rwanda. The specific objectives measured in this study were to translate the LEFS into Kinyarwanda, to test the translated LEFS for clarity, to modify and rectify unclear items found in the LEFS for specific Rwandan cultural activities of daily living (ADL) and thereafter re-establish the intra-and inter-assessor reliability of the modified LEFS.

Methods
The LEFS assesses the subjective functional activity performance of daily living, in the lower extremities. It was developed and validated for a variety of lower extremity conditions based on the WHO model of impairment, disability and handicap (Binkley et al. 1999) particularly for the elderly. The LEFS is expected to measure even small effects of impaired activity performance experienced by participants with lower extremity musculoskeletal dysfunction accurately (Cacchio, De Blasis, Necozione, Rosa, Riddle, di Orio, et al. 2010;Yeung, Wessel, Stratford & Macdermid 2009). The scale assesses the level of difficulty in performing a variety of ADLs. Each activity on the scale is scored by the participant as 0 ¼ 'Extreme difficulty or unable to perform activity', 1 ¼ 'quite a bit of difficulty', 2 ¼ 'Moderate difficulty', 3 ¼ 'A little bit of difficulty' and 4 ¼ 'no difficulty'. The scale scores vary from 0 (none) to 80 (normal) (Binkley et al. 1999).

LEFS translation into Kinyarwanda
The scale was translated from English to Kinyarwanda, by two independent professional language translators from the Language Centre at the College of Medicine and Health Sciences, University of Rwanda (CMHS-UR). Susequently, two independent professional translators translated the scale back to English, to ensure content validity. The translation was assessed by a consensus panel of two physiotherapists and two medical doctors working at the Treatment, Research and AIDS Centre in Rwanda, together with all four translators and the first author of this study. Changes and modifications (indicated in the appendices; Table A2) were made for some scientific terms and functional activities in the scale. The modifications in the activity performance were based on the ADLs that are culturally applicable; an example being a question that asked about having difficulty 'getting in and out of a car'. Most people in Rwanda travel in public taxis/buses (for those who manage to travel in vehicles). When participants were asked about having diificulty getting in and out of a car, some mentioned, 'I have never moved with a car' or 'I seldom move with a car' So, the item/activity 'getting in and out of a car' was modified as having any difficulty of 'getting in and out of a car/public taxi/bus'.

Intra-assessor reliability prior to modification of LEFS
A pilot study was carried out to assess for intra-and inter-assessor reliability of the LEFS. Stage 1 of the study aimed at testing the LEFS for intra-assessor reliability. The translated and contentmodified LEFS was administered to a sample of 50 adults (18 -60 years old) PLHIV on ART, both males and females, who were systematically selected from all the PLHIV on ART registered at an outpatient ART clinic at the Biryogo Health Centre, commonly known as 'Kwa Nyiranuma' in Kigali city. The health centre attends to more than 50 PLHIV on ART on each of the five working days of the week. Ten participants were systematically selected from each list of the first 50 PLHIV attending the centre per day. The selection took from the 5th person and systematically with an interval of 5 up to the 50th person on the list. The sample of 50 participants was obtained in one week. This sample size was the optimal number for feasible pilot study data that are scheduled for only one week. Participants with known deformities and injuries of the lower extremities were excluded from the study. The first author administered the translated LEFS to these selected participants. Two assessments were conducted for each participant, with a week's interval between the two assessments. Participants who could read and write were given the scale to complete with the assessor available for clarification of the scale. Participants who did not know how to read or write had the first author administer the scale by reading each question/item to the participant and recording the responses appropriately. These interviews took place in a private room. Prior to the start of assessing the participants in the pilot study, the assessor had practiced scoring of the scale on five adult PLHIV on ART at the same clinic who were not included in the study. This was done to familiarise the administration and scoring techniques of the scale so as to minimise errors.

Modification of the functional activities in the LEFS
Following the analysis of the intra-assessor correlation between the first and second assessments in stage one, all activities were classified as strong (r ≥ 0.8), moderate (r , 0.8 and ≥ 0.5) and weak (r , 0.5). In addition, during stage one, some activities in the LEFS were unclear to the Rwandan participants and needed precise examples, forming the basis for the subsequent modifications. All such activities were modified and made clearer with specific examples, without changing the concepts and context of the original LEFS. The modification was done in consultation with a team of three health professional experts, two physiotherapists and a medical doctor, who were experienced in rehabilitation services, and two participants. The purpose of the team consultation was to establish appropriate activities that are commonly and culturally undertaken by people living in Rwanda and similar to the activities that define the LEFS. The activities and their common examples were identified.

Intra-and inter-assessor reliability after modification of LEFS
The modified LEFS was then assessed after modifying unclear, moderate and weakly correlated activities in the scale. The intraand inter-assessor reliability was undertaken by three assessors; the first author and two other assessors who were qualified physiotherapists with master's degrees and who were selected by the first author. A sample of 12 participants was randomly selected from both female and male adult PLHIV attending ART clinic at the Kanombe Military Hospital in Kigali, by using random numbers that corresponded to the registration numbering list of the participants at the clinic on one day. Two assessments, one week apart, were carried out to test the intra-and inter-assessor reliability after the above modifications. A two-hour training session was conducted for the two assessors to familiarise them with using the scale. The three assessors administered the scale piloted and modified in stage one. Each assessor carried out the assessment of each participant independently, and was blinded to the other assessors' assessment outcomes and participants' scores.
An ethical clearance certificate (protocol number M080812) for this study was obtained from the Human Research Ethics Committee at the University of the Witwatersrand and the research protocol was approved by the Faculty of Health Sciences at the University. As the research data were collected in Rwanda, national clearance was also obtained from the Institutional Review Board at the College of Medicine and Health Science, University of Rwanda, and scientific approval by the National Commission for control of HIV/AIDS, in Rwanda. Authorisation letters were obtained from the Biryogo Medical Centre and Kanombe Military Hospital where the study was conducted. A letter containing information describing the details of the study was given to the participants to invite them to participate, before they were recruited into the study. Participants, who agreed to participate and gave permission for use of their medical records, signed a consent form. Confidentiality and anonymity were ensured for all participants.
The statistical analysis was done using STATA (version 11, Stata Corp, College Station, TX, USA). The variables (activities) in the LEFS were categorical and ordinal in nature. Spearman's rank correlation coefficient was used to measure statistical independence between the same functional activities at the two assessments done at two intervals for the same participants. The activity correlation coefficients were classified according to the levels of strength, as strongly (r ≥ 0.7), moderately (r , 7 ≥ 0.5) and weakly (r , 0.5) correlated activities. All activities with moderate and weak correlation coefficients (r), according to the classification, in stage one were considered for modification (Table A2).

Discussion
This study represents the first reliability test of the LEFS in patients on ART, from English (Binkley et al. 1999) into Kinyarwanda and adapted for an appropriate cultural context. HIVrelated disability has been associated with decreased physical functioning and has numerous impacts on ADLs (Cacchio et al. 2010;Cade, Peralta & Keyser 2004). The identification of functional activities of the lower extremity is crucial for rehabilitation of patients with chronic illness such as those living with HIV and on ART (Dudgeon, Phillips, Bopp & Hand 2004;O'Brien, Nixon, Tynan & Glazier 2010). This study tested the LEFS to assess the functional activities of lower extremity for rehabilitation purposes in Rwanda. The tested scale can likely be adapted for similar purposes in Africa and other developing countries. The LEFS has very high correlation coefficient (r ¼ 0.94) in the developed world. It was developed and validated for the purpose of identification and evaluation of lower extremity functional activity among the elderly (Binkley et al. 1999). Studies suggest that there might be important differences in health-related activities between highincome and middle/low-income countries (Karlsson, Nilsson, Lyttkens & Leeson 2010). Scales may not identify the activities among the population in a developing environment (Ebrahim & Davey 2001). According to Ebrahim and Davey (2001) research findings from developed settings are not necessarily appropriate to other contexts; thus, local knowledge is important. Our study confirms this, with most of the activities in the original LEFS being only moderate correlations and a few weak. This was probably attributable to the fact that some of the activities in the LEFS were not familiar to most of the population living in Rwanda. In addition, these differences might be reflective of linguistic specificities and cultural differences, but they may also result from methodological disparities such as differences in the clinical characteristics of patients, as reported by Perez, Galvez, Huelbes, Insausti, Bouhassira, Diaz, et al. (2007) in their study which tested the reliability of the Spanish DN4 version from the original French version. It is important that outcome measures used in an environment that is different from the one in which they were originally developed and validated are modified and re-tested. The reliability of the adapted Kinyarwanda version of the LEFS-Modified tool was strong. This implies that the tool can be used by clinicians working at ART clinics, to identify PLHIV with functional limitations at an early stage for appropriate management. Rehabilitation professionals can also use this tool to evaluate progress during rehabilitation. This may improve the quality of care of PLHIV.

Conclusion
Our modified, translated LEFS performed well, with very few remaining moderate and no weak correlations of functional activities in the local environment. Modifications to take into account local conditions are critical for the evaluation of tools that have been validated in developed world contexts. This study modified and re-tested the reliability of the LEFS tool derived from a developed world context, to local conditions in a developing African country. This implies that the modified LEFS can be well used by clinicians, specifically at the ART clinics in Rwanda and possibly other sub-Saharan African countries to screen and identify people with functional limitations at an early stage, for treatment, rehabilitation and or referral to appropriate health-care services, with the aim of improving the QoL of PLHIV. with clinical research training, statistical software package and its application skills and IDRC through APHRC and ADDRF for the financial and scientific support in this study. Professor Venter is supported by PEPFAR.
Tumusiime contributed to study design, data collection and analysis, manuscript preparation and writing. Stewart contributed to research study mentorship, manuscript preparation and writing. Venter contributed to research study mentorship, manuscript preparation and writing. Musenge contributed to data analysis, manuscript preparation and writing.       Table A6. LEFS-Modified. We are interested in knowing whether you are having any difficulty at all with the activities listed below because of your lower limb problem (s). Please provide an answer for each activity. Today, do you or would you have any difficulty with: (Circle one number on each line that corresponds to your appropriate answer)