Development and Validation of the Symptom Assessment to Improve Symptom Control for Institutionalized Elderly Scale

OBJECTIVES: To validate a newly developed multiple symptom self-assessment tool in nursing homes.

DESIGN: Thirty prevalent symptoms identified in the literature were classified by a 2-round Delphi procedure to a top 10 of the most relevant, burdensome symptoms. Because no existing symptom scale fully covered this top 10, we developed a new scale, consisting of a horizontal numerical scale for the top 10 symptoms, with the possibility to add and rate 3 other symptoms. This scale was validated.

SETTING AND PARTICIPANTS: Hundred seventy-four participants, mean age 85 (±5.94) years, were recruited from 7 nursing homes (86%) and 3 acute geriatric wards (14%).

METHODS: To test the construct validity, participants with and without a palliative status were enrolled. Participants completed the Symptom Assessment to Improve Symptom Control for Institutionalized Elderly (SATISFIE) scale on day 0 and day 1 (intrarater reliability). Nurses completed the scale on day 0 (inter-rater reliability). Descriptive statistics described the characteristics of the study population and symptom scores. Differences in symptom scores between palliative and nonpalliative participants were analyzed with the Mann-Whitney U test. Intrarater and inter-rater reliability were calculated by means of an intraclass correlation coefficient. Factor analysis searched for possible symptom clusters. Feasibility was evaluated by measuring the assessment time and by providing a questionnaire for the nurses.

RESULTS: In the nonpalliative group (n = 130), the highest self-rated median scores were pain on day 1 [median 3, interquartile range (IQR) 0-5] and pain on day 2. In the palliative group (n = 44), the highest median self-rated scores were fatigue on day 1 [median 5 (IQR 0-6)], lack of energy on day 1 and 2 [both median 5 (IQR 0-8)]; and depressed feeling on day 2 [median 3 (IQR 0-5)]. Nurse assessments median scores were the highest for depressed feeling [median 5 (IQR 1-7)], fatigue [median 4.5 (IQR 0-6.5)], and lack of energy, [median 3 (IQR 0-6)] in the palliative group. In the nonpalliative group, none of the median scores was 3 or more. Intraclass correlation coefficients for intrarater reliability varied between 0.65 and 0.89 and for inter-rater reliability (patients-nurses) between 0.18 and 0.63. Mean assessment time for nurses was 2.0 minutes [standard deviation (SD) = 1.01]. For participants, it decreased from 10.5 minutes (SD = 5.41) at the first assessment to 7.5 minutes (SD = 3.72) at the second assessment. Nurses determined the SATISFIE instrument to be useful, applicable in daily practice, and sufficiently comprehensible for the patients.

CONCLUSIONS: The SATISFIE scale is a valid and feasible instrument for regular, multiple symptom assessment in institutionalized older persons.