The Connective Tissue Disease Screening Questionnaire (CSQ) is a highly reproducible, sensitive, and moderately specific instrument for screening subjects with potential Connective Tissue Disease (CTD).

Epidemiologic studies of the incidence of connective tissue diseases (CTDs) or their potential associations with exogenous risk factors are limited by the low prevalence, diverse manifestations, lack of specific diagnostic tests, and the expense associated with confirming the diagnosis. Screening by questionnaires such as the CTQ is a useful, validated, efficient case-finding strategy for systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and ankylosing spondylitis.

The CSQ was developed in English and translated to Dutch. It has 30 items and can be completed within 30 minutes.
For all intended purposes a person who does not meet the CTD criteria by the CSQ, as described below, does not have a CTD. The instrument can be used to identify a smaller group for subsequent detailed, specific examination. It should prove useful in population studies of the incidence, prevalence, and etiology of CTDs¹⁾.

An algorithm for identifying potential CTD based on the ACR or published clinical criteria was used.

*Alopecia and Raynaud’s phenomenon were demonstrated by Tan and colleagues to have good sensitivity for SLE but lower specificity in distinguishing SLE from other CTDs. These were included because the goal of our study was to develop a highly sensitive instrument for detecting any CTD.
*Seizures and psychosis were not included because of their relatively low sensitivity (12 and 13%, respectively) in the study by Tan and colleagues. Results of tests for immunologic disorders, anti-Smith (Sm) antibody, anti-double stranded DNA (dsDNA) antibody, and false positive serologic test for syphilis were not ascertained by questionnaire because we thought that most patients would not know the results of these laboratory tests.

See²⁾

Sensitivity and Specificity
Sensitivity of the CSQ (the proportion meeting CTD criteria on the CSQamong those with confirmed disease) ranged from 83% to 96% for individual CTDs. Specificity of the CSQ (the proportion not meeting CTD criteria among those without confirmed disease) ranged from 83 to 93%. Specificity for individual CTDs was better among general medical control subjects than among non-CTD control subjects. Specificity was lowest for polymyositis/dermatomyositis and Sjogren’s syndrome. Likelihood ratios (the ability of the test to distinguish between diseased and nondiseased subjects) ranged from 5.0 for Sjogren’s syndrome to 11.9 for MCTD1.

Stratified Analysis of Specificity
Socioeconomic status was measured by educational level and health insurance status. We found significantly lower specificity among subjects with Medicaid or no health insurance and slightly lower specificity among less-educated subjects and those who were nonwhite. Specificity for detecting certain CTDs was significantly lower in patients with heart disease and depression.

Predictive Values
Positive predictive values in the study population, where the prevalence of CTD was much higher than in the general population, ranged from 23.5 to 84.9%. Positive predictive values in the general population, calculated assuming prevalence rates from population-based studies of CTD, ranged from 0.03% to 9.7%. As one would expect, the highest positive predictive value was for RA, the CTD with the highest prevalence in the general population. Negative predictive values were higher than 92% in the study population and higher than 99% in the general population.

Reliability
The K statistic calculated to evaluate test-retest reliability for detection of any potential CTD was 0.82, a rate indicating “excellent” reproducibility. Test-retest reliability values for specific CTDs were 0.50 for SLE, 0.80 for RA, 0.68 for scleroderma, 0.56 for polymyositis/dermatomyositis, 0.71 for Sjogren’s syndrome, and 0.68 for MCTD.