Desktop version

Home arrow Health arrow Evaluation of health promotion and disease prevention programs : improving population health through evidence-based practice


Evaluation Design and Analysis

An interrupted time series design (TSD) and analyses, based on models for sequential observations over time, evaluated the impact of the MSHA policy. With time-series data, analyses account for the correlation between proximal observations.


Quarterly Fatality Rates

During January 1, 1995, to December 31, 2006, there were 259 fatalities at the study mine population: 160 in the pre-intervention period (O1) and 99 in the post-intervention period (O2). In the 23 quarters in the O1 period, the fatality rate was 0.025 per 200,000 employee hours. In the 25-quarter O2 period, the fatality rate was 0.017, a decline of 33% from the pre-evaluation period. The main intervention effect on the quarterly and semi-annual fatality rates of the MSHA was not statistically significant: (chi-square = 1.12; p = 0.290) and (chi-square = 2.39; p = 0.122). The decline in fatality rates could not be attributed to implementation of MSHA’s Part 46 training regulation.

Serious Injury Rates by Aggregate Type, Ownership, and Production Trends

Data for the 7,998 mines were stratified by relevant BLS codes, by year-round versus intermittent mining operations, and ownership, because 10 large firms were responsible for 50% of annual US surface mine production. Figure 3.2 presents the data comparing serious injury rates at mines, stratified by SIC codes. The period in which MSHA training regulation

Quarterly Rate of Serious Injuries by Aggregate Type

figure 3.2 Quarterly Rate of Serious Injuries by Aggregate Type: 1995-2006

was issued is marked with a dashed vertical line. The effective date of the rule, September 30, 1999, is marked with the solid vertical line. Over the 12-year period, the overall rate of serious injuries declined by 52.6% at crushed stone operations, 46.2% at sand and gravel operations, and 38% at other surface mineral mines. The time series analyses and model identification techniques provided no evidence that the large percent of change documented in serious injury rates in the post-intervention period could be attributed to implementation of the Part 46 training regulation.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >

Related topics