Continuity of Care: Validation of a Self-Report Measure to Assess Client Perceptions of Mental Health Service Delivery

Abstract

The psychometric characteristics of an instrument to assess perceived continuity of care among mental health patients were examined. 441 adults with severe and persistent mental illness were recruited from 70 inpatient, outpatient, emergency and community treatment programs (n = 259 females; M age = 42.5, SD = 10.3 years) in Alberta, Canada. Respondents completed a 43-item self-report questionnaire to rate perceived continuity. Item quality was assessed by examining missing data and frequency distributions. In a randomly selected subsample (n = 171), exploratory factor analysis identified 3 dimensions: (1) perceived attentiveness to individual needs or changes in illness or life circumstances (Individualized Care; alpha = .64); (2) perceptions of a coherent system characterized by good communication between providers (Responsive System; alpha = .71); and (3) perceived responsiveness on the part of a primary provider (Responsive Caregiver; alpha = .52). In a second random subsample (n = 181), confirmatory factor analysis provided support for this 3-factor structure and the inclusion of a second-order "continuity" factor (alpha = .72), and informed further item reduction. Split-half reliability for the second-order factor was calculated. Associations between the factors and "objective" measures of continuity, as well as relevant clinical, quality of life, and service satisfaction variables, are reported. Use of the measure for clinical and research purposes and its limitations are considered.

Measurement Properties of Questionnaires Measuring Continuity of Care: A Systematic Review

“…Seventeen questionnaires measured continuity of care from the perspective of the patien [15]; [16]; [24]–[27]; [29]–[35]; [37]–[41]; [43]–[45], four from the perspective of the care provider/program director [28]; [36]; [42]. From the instruments measuring continuity from the perspective of the patient, three were developed for diabetic patient [29]; [33]; [44], three for patients with a mental illnes [24]; [30]; [37]; [41]; [43], two for patients with cance [38]; [45], two for previously hospitalised patient [26]; [35], two for patients with complex and chronic care need [32]; [40], one for patients with heart failure or atrial fibrillatio [34]; [39], one for users of welfare services [25], one for patients visiting their family practice physician [31], one for patients living at home [27] and one for patients in general regardless of morbidity or care setting [15]; [16].…”

“…Ten instruments measured aspects of personal, team and cross-boundary continuit [15]; [16]; [24]; [26]; [30]–[35]; [37]; [39]; [41]; [44], while eleven instruments measured only one or two of these dimensions [25]; [27]–[29]; [36]; [38]; [40]; [42]; [43]; [45].…”

“…Eight of the instruments were solely developed and/or evaluated in primary care population [27]; [31]–[33]; [40]; [41]; [43]; [44], eight solely in secondary care population [26]; [34]–[36]; [38]; [39]; [42]; [45] and five were developed and/or evaluated in both primary and secondary care populations [15]; [16]; [24]; [25]; [28]–[30]; [37].…”

A tool for assessing continuity of care across care levels: an extended psychometric validation of the CCAENA questionnaire

“…Most available instruments are addressed to specific populations [8], such as patients with diabetes [9,10], cancer [11,12], mental illness [13,14] and patients with an unspecified chronic disease [15] or aimed at users attended to in primary care settings or as inpatients [16,17]. The first generic tool developed to address continuity of care across care levels as perceived by health care users, regardless of morbidity, was the CCAENA© questionnaire (Cuestionario Continuidad Asistencial Entre Niveles de AtenciÓn)[18].…”

Statistical criteria for parallel tests: A comparison of accuracy and power

“…The need for parallel tests arises also on estimation of reliability with the split-half method, which implies partitioning the test at hand into two parallel halves (e.g., Altin & Gençöz, 2009;Joyce et al, 2010;Schmidtke & Metternich, 2009;Woods et al, 2008). Cronbach's alpha is also used rather often for the latter purpose (Hogan, Benjamin, & Brezinski, 2000), but both approaches to reliability estimation have pros and cons (Charter, 2001;Sijtsma, 2009;Thompson, Green, & Yang, 2010).…”