Development and validation of the Patient’s Health Belief Questionnaire on Psychiatric Treatment

Figures & data

Table 1 Patient’s Health Belief Questionnaire on Psychiatric Treatment (N=588)

Items	Mean±SD	Source
I am directly responsible for my condition getting better or worse	4.7±1.7	MHLC-Internal
If I see my doctor regularly, I am less likely to have problems with my condition	4.6±1.7	MHLC-Doctor
When someone forces me to do something, I feel like doing the opposite	2.7±1.8	HPRS
For me, the good things about medication outweigh the bad	4.6±1.7	DAI-10
I feel strange, “doped up”, on medication	3.1±2.0	DAI-10
The main thing which affects my condition is what I myself do	4.2±1.8	MHLC-Internal
Following doctor’s orders to the letter is the best way to keep my condition from getting any worse	4.9±1.5	MHLC-Doctor
I resist the attempts of others to influence me	3.3±1.9	HPRS
Medications make me feel more relaxed	4.9±1.5	DAI-10
Medication makes me feel tired and sluggish	3.7±2.0	DAI-10
I feel more normal on medication	4.3±1.8	DAI-10
If my condition takes a turn for the worse, it is because I have not been taking proper care of myself	4.2±1.9	MHLC-Internal
Whenever my condition worsens, I should consult a medically trained professional	5.5±1.2	MHLC-Doctor
It is unnatural for my mind and body to be controlled by medications	2.9±1.9	DAI-10
My thoughts are clearer on medication	4.0±1.9	DAI-10
Taking medication will prevent me from having a breakdown	4.1±1.9	DAI-10
I become angry when my freedom of choice is restricted	4.2±1.8	HPRS
Subscales
Attitudes toward medication
Positive Aspects of Medications	18.1±4.8	5 items from DAI-10
Negative Aspects of Medications	9.7±4.2	3 items from DAI-10
Perception of health controls
Doctor-HLOC	15.1±3.4	3 items from MLHC-Doctor
Internal-HLOC	12.9±4.2	3 items from MHLC-Internal
Psychological Reactance	10.2±3.8	3 items from HPRS

Abbreviations: DAI-10, Drug Attitude Inventory-10 items; HLOC, health locus of control; HPRS, Hong Psychological Reactance Scale; MHLC-Doctor, Multidimensional Health Locus of Control-Doctor subscale; MHLC-Internal, Multidimensional Health Locus of Control-Internal subscale.

Table 2 Sociodemographic and clinical characteristics of the samples studied (588 psychiatric patients, 1,114 psychiatric drugs used)

Variables	Mean±SD	Percentage
Age (years)	45.7±13.1
Time as psychiatric patient (months)	85.2±93.6
Number of admissionsa	2.64±2.66
Number of psychiatric medications	1.9±1.1
Duration of psychiatric medication use (months)	42.0±50.2
Sex
Female		53% (310/588)
Male		47% (278/588)
Educational level
Can read and write		3% (16/588)
Primary school		27% (160/588)
Secondary school		44% (258/588)
University		26% (154/588)
Diagnosis
Depressive disorder		38% (224/588)
Anxiety disorder		28% (169/588)
Schizophrenia		23% (134/588)
Bipolar disorder		7% (40/588)
Personality disorder		3% (16/588)
Substance use disorder		<1% (1/588)
Other diagnoses		<1% (4/588)
Prior psychiatric admission		28% (165/588)
Prior involuntary admission		19% (111/588)
Self-reported adherence
High		44% (490/588)
Good		7% (81/588)
Moderate		15% (165/588)
Poor		11% (128/588)
Very poor		7% (73/588)
Discontinuation		16% (177/588)

Note:

a In 156 patients with at least one admission.

Table 3 Odds ratios (95% confidence interval) of five continuous subscales (dependent variable: adherence [yes/no]) (N=1,114 drugs)

Subscales	Univariate	Multivariate (only 5 subscales)a
Positive Aspects of Medications	1.07 (1.04–1.10)	1.39 (1.01–1.07)
Negative Aspects of Medications	0.917 (0.889–0.945)	0.938 (0.909–0.969)
Doctor-HLOC	1.12 (1.08–1.16)	0.961 (0.929–0.993)
Internal-HLOC	0.990 (0.961–1.02)	1.10 (1.06–1.15)
Psychological Reactance	0.945 (0.914–978)	0.963 (0.929–0.998)

Notes:

a Hosmer and Lemeshow test: chi-square 13.497, df 8, P=0.096. None of the confounding variables (gender, age, and education) were significant and were not included in the backward step model.

Abbreviation: HLOC, health locus of control.

Table 4 Odds ratios (95% confidence interval) for five dichotomized psychological subscales (dependent variable: adherence [yes/no]) (N=1,114 drugs)

Subscales	Univariate	Multivariate (only 5 subscales)a	Multivariate (confounders)b
High Positive Aspects of Medications	1.48 (1.16–1.91)	1.28 (0.98–1.68)	ns
Low Negative Aspects of Medications	1.89 (1.47–2.43)	1.70 (1.32–2.20)	1.76 (1.36–2.29)
High Doctor-HLOC	1.77 (1.37–2.28)	1.57 (1.20–2.05)	1.61 (1.24–2.09)
Low Internal-HLOC	1.16 (0.90–1.49)	1.25 (0.96–1.61)	ns
Low Psychological Reactance	1.44 (1.12–1.85)	1.25 (0.97–1.62)	1.31 (1.01–1.70)

Notes:

a Hosmer and Lemeshow test: chi-square 3.546, df 8, P=0.896.

b Confounding variables include education in years (P=0.073). Gender and age were not significant and were not included in the backward step model. Hosmer and Lemeshow test: chi-square 8.211, df 8, P=0.413.

Abbreviations: HLOC, health locus of control; ns, not significant.

Table 5 Summary of CHAID analysis group

Subgroup of patients	N	Adequate adherence (%)	Node
Low Negative Aspects of Medications and high Psychological Reactance	259	67	4
Low Negative Aspects of Medications, low Psychological Reactance, and high Internal-HLOC	157	69	8
Low Negative Aspects of Medications, low Psychological Reactance, and low Internal-HLOC	204	82	7
High Negative Aspects of Medications and low Doctor-HLOC	303	50	6
High Negative Aspects of Medications and high Doctor-HLOC	303	71	5

Abbreviations: CHAID, Chi-squared Automatic Interaction Detector; HLOC, health locus of control.

Table 6 Correlations among the five subscales (N=588 individuals)

	Negative Aspects of Medication	Doctor-HLOC	Internal-HLOC	Psychological Reactance
Positive Aspects of Medications	−0.19	0.38	0.20	−0.15
Negative Aspects of Medications	–	−0.10	−0.19	0.22
Doctor-HLOC	–	–	0.22	−0.12
Internal-HLOC	–	–	–	0.09

Abbreviation: HLOC, health locus of control.