RESUMO
Effective doctor-patient-family communication is an integral and sensitive part of health care, assessing its quality is essential to identify aspects needing disclosure and, if necessary, improvement. Cross-sectional study aimed to analyze the sources of evidence of validity and the number of participants needed to reliably apply the Quality of Communication Questionnaire (QoC) through Generalizability Theory (GT). The mean age of the 150 patients hospitalized at the end of life was 50.5 (SD = 13.8) years, the mean hospital length of stay was 7.5 (SD = 10.2) days, 56.9% were male. Regarding the 105 patients' family members of patients whose mean length of hospital stay was 9.5 (SD = 9.1) days, their mean age was 42.2 (SD = 14.7) years, 69.5% were female. GT was used to quantify the minimum number of questionnaires needed, with the aim of reaching a reliable estimate of QoC with G-coefficients. To reach a reliability of .90, there is a need for 25 for the Eρ2 questionnaires and 35 for the Φ. The exact estimation identified the minimum number of questionnaires required for the evaluation of physicians by patients. To obtain a reliability of .90, there is a need for 30 and 40 questionnaires for the G-coefficients. A practical and fast application makes it possible to use QoC in its entirety or alone to evaluate general communication or communication about palliative care. Furthermore, based on these results, it was possible to identify which aspects were effective or ineffective in these contexts.
RESUMO
Assessment of communication between physicians and patients' family members is essential to improving healthcare quality. To adapt the Quality of Communication Questionnaire (QoC) for family members and to analyze its validity evidence for use in Brazil. Data were collected between 2017 and 2019, with family members of patients in intensive care (IC) and palliative care (PC) from five public hospitals in the South Brazil. The QoC was adapted for family members for use in Brazil, and its cross-cultural adaptation was carried out. The clarity and cultural appropriateness of the pre-final version were evaluated by 30 family members of patients in IC. The final version was responded by 198 family members of patients. All items were considered clear, and appropriate to Brazilian culture. The goodness of fit index for proposed model had CFI 0.96 (CI95%: 0.94 - 0.98), TLI 0.95 (CI95%: 0.92 - 0.97), RMSEA 0.07 (CI90%: 0.06 - 0.08), and χ2/df 2.18. Cronbach's alpha coefficient (α) among family members of patients in PC was 0.88 for the general communication (first subscale) and 0.80 for the end-of-life communication (second subscale). However, among family members of patients in IC, α was 0.86 for the first subscale and only 0.53 for the second subscale. The QoC for family members and its cross-cultural adaptation were carried out successfully. It has strong validity evidence among those with loved ones in PC, but only the QoC general communication subscale has strong validity evidence among those with loved ones in IC.
