Related Experiment Video
Updated: Feb 9, 2026
Hydra, a Computer-Based Platform for Aiding Clinicians in Cardiovascular Analysis and Diagnosis
Published on: September 26, 2018
[Prevalence of Cardiovascular Risk Factors at The Population Level: A Comparison of Ambulatory Physician-Coded Claims
Aniela Angelow1, Katrin Christiane Reber2, Carsten Oliver Schmidt3
1Abt. Allgemeinmedizin, Institut für Community Medicine, Universitätsmedizin Greifswald, Greifswald.
Insights
The validity of ICD-10 codes for cardiovascular risk factors in claims data varies significantly. While diabetes and hypertension showed higher agreement, smoking, obesity, and dyslipidemia were frequently undercoded, impacting prevalence estimates.
Area of Science:
- Epidemiology
- Health Services Research
- Medical Informatics
Background:
- Claims data are widely used for epidemiological research.
- Accurate coding of cardiovascular risk factors (arterial hypertension, diabetes, dyslipidemia, smoking, obesity) in claims data is crucial for reliable health statistics.
- The validity of International Classification of Diseases, 10th Revision (ICD-10) codes in claims data compared to gold-standard clinical measurements is often uncertain.
Purpose of the Study:
- To assess the accuracy of ICD-10 coded cardiovascular risk factors in health insurance claims data.
- To compare the validity of ICD-10 codes for arterial hypertension, diabetes, dyslipidemia, smoking, and obesity against gold-standard clinical data.
- To evaluate the impact of coding practices on population-based prevalence estimates.
Main Methods:
- Utilized data from 1941 participants in the Study of Health in Pomerania (SHIP).
- Linked SHIP's gold-standard clinical data (2008-2012) with electronic medical records from statutory health insurance physicians for billing purposes.
- Assessed agreement between ICD-10 codes (I10.-, E10-E14, E78.-, F17.-, E65-E68) and clinical diagnoses for key cardiovascular risk factors.
Main Results:
- High agreement was observed for diabetes (sensitivity 84%, specificity 95%, PPV 80%) and hypertension (sensitivity 72%, specificity 93%, PPV 97%).
- Low agreement was found for smoking (sensitivity 18%, specificity 99%, PPV 89%), obesity (sensitivity 22%, specificity 99%, PPV 99%), and dyslipidemia (sensitivity 40%, specificity 60%, PPV 70%).
- Factors such as medication, comorbidities, age, sex, and clinical severity influenced ICD-code agreement.
Conclusions:
- The quality of ICD-10 coding for cardiovascular risk factors in ambulatory care is highly variable.
- Diagnoses were generally undercoded, with higher frequency for codes relevant to billing.
- Results highlight potential for errors in population-based prevalence estimates derived from claims data and inform data interpretation.
Objective:
The study assesses the validity of ICD-10 coded cardiovascular risk factors in claims data using gold-standard measurements from a population-based study for arterial hypertension, diabetes, dyslipidemia, smoking and obesity as a reference.
Methods:
Data of 1941 participants (46 % male, mean age 58±13 years) of the Study of Health in Pomerania (SHIP) were linked to electronic medical records from the regional association of statutory health insurance physicians from 2008 to 2012 used for billing purposes. Clinical data from SHIP was used as a gold standard to assess the agreement with claims data for ICD-10 codes I10.- (arterial hypertension), E10.- to E14.- (diabetes mellitus), E78.- (dyslipidemia), F17.- (smoking) and E65.- to E68.- (obesity).
Results:
A higher agreement between ICD-coded and clinical diagnosis was found for diabetes (sensitivity (sens) 84%, specificity (spec) 95%, positive predictive value (ppv) 80%) and hypertension (sens 72%, spec 93%, ppv 97%) and a low level of agreement for smoking (sens 18%, spec 99%, ppv 89%), obesity (sens 22%, spec 99%, ppv 99%) and dyslipidemia (sens 40%, spec 60%, ppv 70%). Depending on the investigated cardiovascular risk factor, medication, documented additional cardiovascular co-morbidities, age, sex and clinical severity were associated with the ICD-coded cardiovascular risk factor.
Conclusion:
The quality of ICD-coding in ambulatory care is highly variable for different cardiovascular risk factors and outcomes. Diagnoses were generally undercoded, but those relevant for billing were coded more frequently. Our results can be used to quantify errors in population-based estimates of prevalence based on claims data for the investigated cardiovascular risk factors.
Related Concept Videos
01:12Analysis of Population Pharmacokinetic Data
01:11Testing a Claim about Mean: Known Population SD
Estimating a population mean requires the samples to be distributed normally. The data should be collected from the randomly selected samples having no sampling bias. The sample size needed to be higher than 30, and most importantly, the population standard deviation should be already known.
In most realistic situations, the population standard deviation is often unknown, but in rare circumstances, when it...
01:24Testing a Claim about Population Proportion
There are two methods of testing a claim about a population proportion: (1) Using the sample proportion from the data where a binomial distribution is approximated to the normal distribution and (2) Using the binomial probabilities calculated from the data.
The first method uses normal distribution as an approximation to the binomial distribution. The requirements are as follows: sample size is large...
01:21Testing a Claim about Mean: Unknown Population SD
Estimating a population mean requires the samples to be approximately normally distributed. The data should be collected from the randomly selected samples having no sampling bias. There is no specific requirement for sample size. But if the sample size is less than 30, and we don't know the population standard deviation, a different approach is used;...
01:25What is Population Genetics?
02:04Conservation of Small Populations