Phase I: a cross-sectional study
Instruments
Depressive symptom screening questionnaire
A standard screening questionnaire by the Centre for Epidemiologic Studies-Depression Scale (CES-D) Thai version was used since it is widely employed for screening adolescents having depressed mood. The questionnaire was validated showing high internal consistency with Cronbach’s alpha coefficients 0.86. Sensitivity, specificity and accuracy of this version were 72, 85 and 82 %, respectively (Trangkasombat et al. 1997). The questionnaire contains 20 four-scale questions (rarely = 0, occasionally = 1, some = 2, most = 3). The total score was computed by summation of the score given for each question, thus ranging from 0 to 60. It is recommended that individuals with the CES-D score ≥22 were potentially having depressive mood, so-called ‘depressive student’ (Trangkasombat et al. 1997). A form to record the student profile was also enclosed as we wished to identify demographic factors relating to depressed mood. Demographic factors were gender, age, faculty and year of study which were correlated to depressive symptom stated previously (Othman et al. 2013; Shamsuddin et al. 2013). Grade Point Average (GPA), used as a key indicator for study performance, was also recorded because academic requirement was proved to be a major stressor for students (Othman et al. 2013). Name and telephone number were requested only to further contact the participant regarding the experimental study.
Suicidal screening questionnaire
Participants with CES-D score ≥22 were asked to complete a suicidal screening questionnaire because depressive symptom may lead to suicide (Yongthong et al. 2013). This questionnaire was developed by the Department of Mental Health, Ministry of Public Health of Thailand (2005) (Ministry of Public Health 2014). It was specific to assess suicidal ideation of participants in the past 2 weeks. This questionnaire contains 10 dichotomous questions (yes/no). The answer of ‘yes’ to at least one question means that participant was at risk of suicide. In this step, all participants with suicidal ideation were referred to the hospital.
Screening procedure
Depressive symptom screening procedure was conducted in July 2010. All students from four health science faculties of MSU (nursing, medicine, pharmacy and public health) were invited to self-complete the CES-D questionnaire. Those with CES-D score ≥22 were asked to complete the suicidal screening subsequently. If they were already diagnosed with depression, they were then excluded from the study. If they were at risk of suicide, they were then referred to the hospital.
Statistical analysis
Multiple linear regression was performed to identify potential predictors of depressive symptom by using CES-D score as a dependent variable. Independent variables including gender, age, faculty, year of study and GPA were entered in the analysis by stepwise method. Stepping method criteria used P value of F which was set at 0.05 for entry and 0.10 for removal.
Phase II: experimental study
Study design
A randomized controlled trial was conducted at a community pharmacy located on the university campus. The trial was undertaken from August to December of 2010.
Participants
The target population was depressive students, found in phase I. Optimal sample size was calculated using the formula below. Parameters used for calculation were based on statistical standard practice (Zα = 1.96) and previous research, which reported the change of depressed outcomes due to pharmacist’s intervention (σ2 = 6.22, E = 2.26) (Wisethorn 2008). This indicated that the appropriate number of participants should be at least 25 in each group, but in this study that number was increased by 30 % in order to account for loss of follow up. Therefore, the appropriate number of participants was 32. All depressive students were invited to the study; however, 68 enrolled into the trial eventually.
$${\text{n}} = {\text{NZ}}^{ 2}_{\alpha } \sigma^{ 2} /({\text{NE}}^{ 2} + {\text{ Z}}^{ 2} \sigma^{ 2} )$$
Randomization
Participants were assigned to either group 1 or group 2 by stratified random sampling. To ensure demographic parity of the two groups, they were allocated based on faculty, year of study and gender.
Intervention
Two community pharmacists were trained by the psychiatric pharmacist for two-weeks at a neurological hospital prior to the study. They practiced giving advice to a mock client. A flowchart describing the interventions used in the experimental study is shown as Fig. 1.
Participants in group 1 (n = 34) received group counselling regarding depressive symptom by a trained community pharmacist. They were scheduled to visit the community pharmacy on a pre-determined day that enabled them to attend the group counselling together. The group session lasted approximately 1 h, covering information about definition and cause of depressive symptom, risk factors, signs, and types of depression, symptoms, treatments, and self-managements. The group counseling sessions were provided at week 0, 4, 8 and 16. Outcomes were measured as described below.
Participants in group 2 (n = 34) received individual counselling from one of the two trained pharmacists. Participants were scheduled to visit the community pharmacy. At week 0, the depression management handbook was provided to all participants for self-study. The handbook developed by a research team was comprised of three parts (a) an overview about depressive symptom including definition, causes, risk factors of depressive symptom and self-managements, (b) a patient profile recording personal information and clinical symptoms, and (c) a form for patients to record their daily activities to relieve depressive symptoms. Participants received individual counseling at week 0, 4, 8, 16 from the same pharmacist. The counseling sessions lasted approximately 1 h and focused on finding the cause of depressive symptom and method to relieve depressed mood. Outcomes were measured as described below.
Outcome measurement
Depressed mood level
Depressed mood level was a key outcome measured by the CES-D questionnaire. Assessment was carried out at week 0 before the counselling started and at week 16 after the counselling was completed. Mean CES-D score was calculated for both groups.
Quality of life (QOL)
Previous studies reported that mental health problems, as in the case of other chronic diseases, generally affects quality-of-life (QOL) (Rapaport et al. 2005; Kanjanasilp et al. 2009), and pharmacist intervention was shown to impact QOL of primary care patients with depression (Rubio-Valera et al. 2013). This study, therefore, evaluated and compared QOL before and after the pharmacist intervention. The Short Form (36) Health Survey (SF-36) Thai version was used to measure QOL. Lim et al. (2008) validated this instrument thoroughly and found good reliability of each domain test, presenting the Cronbach’s alpha coefficient greater than 0.70. SF-36 consisted of 36 items clustered to measure eight domains of QOL: physical functioning (PF), role-physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role-emotional (RE), and mental health (MH). Details of scales and scoring were explained by Lim et al. (2008). We calculated the mean score for physical health by the summed score of PF, RP, PB and GH divided by four. Similarly, we then calculated the mean score for mental health by the summed score of VT, SF, RE and MH divided by four.
Statistical analysis
Statistical analysis was performed using the statistical package for the social science (SPSS) version 16. All analyses used the intention-to-treat (ITT) approach with a use of sequential hot deck imputation to manage missing data. Descriptive statistics to report results included mean, standard deviation, and percentage as appropriate. Statistical differences within the same group were tested using paired t-test while an independent t-test was used for testing the differences between two groups. The level of statistical significance was determined to be 0.05 or less.
Ethical consideration
The study protocol was approved by Mahasarakham University Ethics Committee. Written consent form was obtained. Information obtained from participants was anonymous and kept confidential.