Interpreting Health Data and Trends

A calculation says what the number is. Interpretation says what that number means for the question.

7.4.1 Separate observation from interpretation

Observation: recorded attendance declined from 200 in January to 150 in March. Interpretation: the decline may warrant investigation. What we cannot conclude: these figures do not establish why attendance declined. Service availability, demand, staffing, recording practice and reporting completeness are all possible explanations.

7.4.2 Calculate absolute change

Absolute change is the new value minus the old value. 150 − 200 = −50. Recorded attendance decreased by 50 visits.

7.4.3 Calculate percentage change

Percentage change is (new − old) ÷ old × 100. (150 − 200) ÷ 200 × 100 = −25%. Recorded attendance declined by 25%. The minus sign means a decrease.

7.4.4 Avoid misleading comparisons

A large hospital with 1,000 recorded visits and a small clinic with 200 are not automatically ranked by performance. You may need the population served, the services available, the operating hours, reporting completeness, and whether the question is about volume, coverage, efficiency or something else.

7.4.5 Check the denominator

Facility A received 80 of 100 expected reports: 80%. Facility B received 45 of 50: 90%. Facility A received more reports. Facility B had higher completeness. Counts and percentages answer different questions.

7.4.6 Consider the reporting period

Do not compare one month with three months as if they were the same window. Confirm the periods, definitions, geography, population or reporting units, methods, and whether revisions or late submissions changed the file.

7.4.7 Use evidence to formulate recommendations

If completeness fell from 90% to 70%, a reasonable next step is to review the missing reports and investigate the decline before the performance assessment is final. “Dismiss the reporting officers because they are not working hard enough” assigns blame without a cause.