Test-retest reliability of the Social Problem Questionnaire in primary care in Italy

The 1-week test-retest reliability of the Social Problem Questionnaire (SPQ) was investigated in a sample of 100 primary care attenders in Italy. The questionnaire was found to be simple and readily acceptable to primary care attenders. Reliability coefficients were high for housing, financial and work problems (Pearson's r and the intraclass correlation coefficient > or = 0.75) and relatively low for problems with spouse or partner (Pearson's r and the intraclass correlation coefficient = 0.30). Pearson's r and the intraclass correlation coefficient were 0.77 and 0.76, respectively, for the overall severity score reported on the SPQ; the two coefficients were 0.63 and 0.62, respectively, for the overall number of problems reported on the SPQ. When the proportion of specific agreement was computed, agreement on absence of social problems was higher compared to agreement on presence. Reliability coefficients tended to be higher in males compared to females. Systematic differences between scores on the first and second tests were found for problems in housing conditions and social relationships excluding relatives (lower scores on the second test), as well as for the residual category "other social problems" (higher scores on the second test).