DISCUSSION:
The present study provides evidence of reliability and validity testing of Mal-PCOSQ. The present study shows that Malaysian women with PCOS are affected by their condition as evidence by low scores firstly in infertility and then followed by weight domain. Body hair has highest score and was the least concern in the local population. Present study was more similar to that of Huang-et-al[2], which is of similar Asian population. Huang-et-al[2] had showed that the burden of hirsutism on HRQoL in South Asians was especially to a lesser extent than that in Caucasians, with weight being the worst domain, followed by infertility using the validated Chinese version of PCOSQ. Cronin et-al[6] found that emotions has lowest score however in contrast to present study, infertility subscales were the second most affected. It also has the highest impact score among African-American women with PCOS compare to the white caucasians. On the other hand, the Iranian study[7] findings showed that Iranians patients were more affected firstly by infertility and then menstruation. Interesting to see, regardless of ethnicity, infertility seems to affect patients with PCOS tremendously regardless of ethnicity.[13] However the second most affected domains varies among different populations which may be due to different cultures and social expectations. This is may be due to the fact that these PCOS patients were recruited in gynaecology clinic, hence infertility is the most common clinical complaint among them.
There was no dropout in our study as the investigator prior to collecting the forms checked the questionnaires. None of the patients has difficulties in understanding the questions in Malay hence indicating that Mal-PCOSQ was feasible to be use in local population. All the coefficients of the Mal-PCOSQ domains are analysed using test-retest reliability analysis which proves a good reliability with Croncbach alpha >0.7; with average of 0.852.
The floor and ceiling effects were small in all subscales of Mal-PCOSQ when compared with SF-36v2. These indicate that the instrument was more specific towards certain areas pertaining to the disease or condition itself, which was better than the generic instrument.
The SF 36v2 questionnaire analysis showed both mental and physical score summary are low, with former 44.83 and later 50.15 which showed that our PCOS patients are significantly affected in both mental and physical health which supported the findings of Mal-PCOSQ which again enhanced the validity of Mal-PCOSQ. The convergent validity of the Mal-PCOSQ was proven by its correlation with similar domains in SF-36v2 whereby the ‘emotion’ domain (Mal-PCOSQ) correlated well with the ‘role emotion’ and ‘mental health’ of SF-36v2. Known group comparisons which were obesity, hirsutism, irregular menses and subfertility indicate that Mal-PCOSQ were able to differentiate between subgroups with different clinical features hence proved its discriminant validity of the instrument.
Infertility and weight appear to be the most affected components in our study population. Patient who has infertility score the domain on Mal-PCOSQ lowest (p=0.006) while obese patients also has significant impact on weight domain on Mal-PCOSQ (p=0.000).(Table 5)
The Mal-PCOSQ can potentially be used upon the diagnosis of the disease and during follow-ups to ascertain which component that affect the patient most and assess the efficacy of the treatment provided. No doubt, this will further improve patients care and outcome as each patient’s management is tailored management to cater for her needs.[12]