Psychometric Properties of the Confidence and Trust in Delivery Questionnaire (CTDQ)

A Pilot Study

Elke Jeschke; Thomas Ostermann; Natalie Dippong; Dagmar Brauer; Harald Matthes

Disclosures

BMC Womens Health. 2012;12(26) 

In This Article

Results

Study Population

The mean age of participants was 31.9 years (SD 4.9; range 19–45). 177 (80.1%) of them attended a childbirth preparation class and 125 (56.6%) were expecting a child for the first time. Complete demographic and socio-medical information on all participants is provided in Table 2.

Dimensions and Internal Reliability of CTDQ

We found a 4-factor model with a Kaiser-Meyer-Olkin measure of sampling adequacy of 0.789 and a highly significant Bartlett test of sphericity (p < 0.001). The cumulative variance explained by this model was 69.6%. The internal consistency of the whole item pool (Cronbach's α = 0.789) and the respective subscales (Cronbach's α between 0.595 and 0.824) can be regarded as sufficient or even excellent. Both factor structure and reliability parameters are shown in Table 1.

The first factor with 4 items describes "confidence in labor (CL)" with items such as "upcoming labor is a source of joy". Its internal consistency is excellent (Cronbach's α =0.824); item-difficulty values between 0.35 and 0.49 can be regarded as excellent (Table 1). Factor 2 consists of 3 items dealing with "partner's support (PS)". Internal consistency (Cronbach's α =0.619) was sufficient, while the item-difficulty (between 0.19 and 0.22) indicates strong floor effects. The next factor with 2 items shortly describes "trust in medical competency (TMC)" with a reliability of α =0.675. Item-difficulty again was quite low, ranging from 0.23 to 0.30. The last factor describes the feeling of "being informed (BI)" about labor and is comprised of one general item ("I feel well prepared for my labor") and one special item on EA ("I feel well informed about EA"). This factor shows the lowest reliability (Cronbach's α =0.595) with an acceptable item-difficulty of around 0.37.

As seen in Table 3, only the "partner's support" correlated strongly with "trust in medical competence" (r = 0.702) while the other subscales showed moderate correlations between r = 0.184 and r = 0.334. With values between 0.213 and 0.632 for the complete instrument and values between 0.365 and 0.717 for the subscales, item-total correlations are in the optimal range for psychological test instruments.[17]

External Validity

To assess external validity, we correlated the CTDQ subscales with the SOC subscales (Table 3). We only observed some minor respective moderate correlations between r = −0.142 and r = −0.325 in absolute values. The strongest correlation was found between the SOC subscale "Manageability" with the CDTQ subscales "Confidence in labor" (r = −0.325) and Partner's support (r = −0.321). Thus the percentages of shared variance indicated by the square of correlation coefficient r2 between the SOC and the CTDQ lies between 2% and 11% and indicates an independence of the scales.

Strong correlations however were recognized between the SOC subscales: all subscales were highly dependent with absolute correlation values ranging from r = 0.506 to r = 0.907. Thus, shared variance lies between 26% and 82% and underpins a interdependence of the SOC subscales. The highest correlations within the CDTQ was observed between the subscales "Trust in medical competence" and "Partner's support (r = 0.702; r2 = 49%), while other subscales correlations ranged between r = 0,184 and 0,334 leading to a shared variance between 3% and 11%. A detailed overview is provided in Table 3.

Subscales and Parity

Scales both of the SOC and the CDTQ did not significantly differ with respect to parity status, age and educational level of the upcoming mothers. In all subgroup analyses score values showed a quite similar range and differences did not turn out to be significant between the groups (Table 4).

processing....