Page 17 - NHMS2015-tradisional and complementary medicine volume IV

Traditional and Complementary Medicine (T&CM)
VOLUME IV

The sample size calculation based on criteria below;

1. Expected T&CM used based on previous study: Prevalence of ever use in a
lifetime or current use within the last twelve month

2. Margin of Error (e) between 0.01-0.05
3. 95% Confidence Interval

To ensure optimum sample size, a few adjustments were made;

1. Adjusted n (srs) for the total number of target population (N) (Based on estimated
2014 population).

n SRS

2. Adjusted for design effect (deff) (Based on previous survey: NHMS 2011)

N (complex) = n * deff

3. Adjusted the n(complex) taking into account expected non response rates of 35%
n (adj) = n (complex) * (1 + non response)

Urban Rural

Thus, the optimum sample size required was 10,428 living quarters.

Sampling Design

The Two Stage Stratified Random Sampling design was used to ensure the national

representativeness for this study. Sampling involved two stages; the Primary Sampling
Unit (PSU), which is EBs and Secondary Sampling Unit (SSU) which is LQ within the

selected EBs. A total of 794 EBs were selected from the total EBs in Malaysia, which

484 and 310 EBs were selected from urban and rural area respectively. Twelve LQs
were randomly selected from each selected EBs. Only one household member within

the selected household was randomly chosen using modified Kish Grid Table (6) to

answer the T&CM module.

15
NATIONAL HEALTH AND MORBIDITY SURVEY 2015 MINISTRY OF HEALTH MALAYSIA