Table 1 Data Quality Measures
From: A basic model for assessing primary health care electronic medical record data quality
Domain | Measure | Operationalization of Measure |
|---|---|---|
Comparability | Comparison of Database Population to a Standard Population | ∙ Compare age-sex structure of the database population to an external/standard population (e.g. population census) using 5 year age bands, graph results ∘ Statistical test: Chi Square by patient age and sex ∙ Compute the mean and median age by sex for the database population and compare to an external/standard population |
Concordance of Test Conditions | ∙ Compute crude and age standardized prevalence of test conditions (diabetes mellitus, hypothyroidism, hypertension, asthma, urinary tract infection, and obesity) within the database population, compare to published prevalence figures for the test conditions | |
Completeness | Sensitivity | ∙ Calculate sensitivity values for test conditions of diabetes mellitus, hypertension, hypothyroidism, asthma, obesity and urinary tract infection. Use test condition definitions as the gold standard and billing code as the comparison standard. |
“Consistency of Capture” [46] | ∙ Calculate: ∙ Percentage of all patients with 1 or more entries on problem list ∙ Percentage of all patients with 1 or more entries on allergy record, including “no allergies” ∙ Percentage of patients visiting in the last year of the database with 1 or more prescribed medications | |
Recording of Blood Pressure, Height, and Weight | ∙ Calculate proportion of patients with: ∙ 1 or more blood pressure recordings for patients 18 + years, males and females ∙ 1 or more height recordings for patients of all ages, males and females ∙ 1 or more weight recording for patients of all ages, males and females ∙ Statistical test: Chi Square by patient age and sex | |
Recording of Blood Pressure among Patients Requiring a Blood Pressure Measurement | ∙ Calculate: ∙ Percentage of patients with diabetes mellitus, with 1 or more blood pressure recordings within 12 months of date of onset of diabetes ∙ Percentage of patients with hypertension medications (2 or more oral anti-hypertensives, or 1 or more diuretics) with 1 or more blood pressure recordings within 12 months of date of onset of hypertension | |
Correctness | Positive Predictive Value | ∙ Calculate positive predictive values (using same approach as for sensitivity) for test conditions of diabetes mellitus, hypertension, hypothyroidism, asthma, obesity and urinary tract infection |
Unlikely Combinations of Age & Specific Procedures | ∙ Calculate the percentage of patients 10 or more years of age with a tetanus toxoid conjugate vaccine (diphtheria, haemophilus B, pertussis, polio, and tetanus) (which is usually reserved for children < 10 years of age) | |
Currency | Timeliness of Weight Recording for Patients with Obesity | ∙ Calculate the percentage of obese patients with 1 or more weight recordings within 1 year of last visit in recorded in the database |
Timeliness of Visit for Pregnancy | ∙ Calculate percentage of patients with a positive pregnancy laboratory test result and 1 or more visits within two months of the result | |
Timeliness of Blood Pressure, Height, and Weight Recording | Calculate proportion of patients with values recorded no more than one year prior to their last visit in the database for: ∙ 1 or more blood pressure recordings for patients 18 + years, males and females ∙ 1 or more height recordings for patients of all ages, males and females ∙ 1 or more weight recording for patients of all ages, males and females Statistical test: Chi Square by patient age and sex |