External Validation of EPICON: A Grouping System for Estimating Morbidity Rates Using Electronic Medical Records

Department of Medical Informatics, Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands.
Journal of the American Medical Informatics Association (Impact Factor: 3.5). 09/2008; 15(6):770-5. DOI: 10.1197/jamia.M2774
Source: DBLP


To externally validate EPICON, a computerized system for grouping diagnoses from EMRs in general practice into episodes of care. These episodes can be used for estimating morbidity rates.
Comparative observational study.
Morbidity rates from an independent dataset, based on episode-oriented EMRs, were used as the gold standard. The EMRs in this dataset contained diagnoses which were manually grouped by GPs. The authors ungrouped these diagnoses and regrouped them automatically into episodes using EPICON. The authors then used these episodes to estimate morbidity rates that were compared to the gold standard. The differences between the two sets of morbidity rates were calculated and the authors analyzed large as well as structural differences to establish possible causes.
In general, the morbidity rates based on EPICON deviate only slightly from the gold standard. Out of 675 diagnoses, 36 (5%) were considered to be deviating diagnoses. The deviating diagnoses showed differences for two main reasons: "differences in rules between the two methods of episode construction" and "inadequate performance of EPICON."
The EPICON system performs well for the large majority of the morbidity rates. We can therefore conclude that EPICON is useful for grouping episodes to estimate morbidity rates using EMRs from general practices. Morbidity rates of diseases with a broad range of symptoms should, however, be interpreted cautiously.

Download full-text


Available from: Robert A Verheij,
  • Source
    • "This primary care-based network was established in 1989, and consists of three group practices in the northern part of the Netherlands with about 20 GPs and about 30,000 patients [16]. The RNG is a validated register [17]. All contacts, diagnoses, referrals and prescriptions are registered in the RNG using the International Classification of Primary Care (ICPC) [18]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Earlier research showed that healthcare in stroke could be better organized, aiming for improved survival and less comorbidity. Therefore, in 2004 the Dutch College of General Practitioners (NHG) and the Dutch Association of Neurology (NVN) introduced the ‘Dutch Transmural Protocol TIA/CVA’ (the LTA) to improve survival, minimize the risk of stroke recurrence, and increase quality of life after stroke. This study examines whether survival improved after implementation of the new protocol, and whether there was an increase in contacts with the general practitioner (GP)/nurse practitioner, registration of comorbidity and prescription of medication. Methods From the primary care database of the Registration Network Groningen (RNG) two cohorts were composed: one cohort compiled before and one after introduction of the LTA. Cohort 1 (n = 131, first stroke 2001–2002) was compared with cohort 2 (n = 132, first stroke 2005–2006) with regard to survival and the secondary outcomes. Results Comparison of the two cohorts showed no significant improvement in survival. In cohort 2, the number of contacts with the GP was significantly lower and with the nurse practitioner significantly higher, compared with cohort 1. All risk factors for stroke were more prevalent in cohort 2, but were only significant for hypercholesterolemia. In both cohorts more medication was prescribed after stroke, whereas ACE inhibitors were prescribed more frequently only in cohort 2. Conclusion No major changes in survival and secondary outcomes were apparent after introduction of the LTA. Although, there was a small improvement in secondary prevention, this study shows that optimal treatment after introduction of the LTA has not yet been achieved.
    BMC Family Practice 06/2013; 14(1):74. DOI:10.1186/1471-2296-14-74 · 1.67 Impact Factor
  • Source
    • "The RNG register is a validated [10] and structured register with regular meetings of participating general practices twice a year, for purposes of maintaining an unambiguous registration, which is an ongoing process. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Previous research on time to referral to orthopaedic surgery has predominantly used hip complaints as starting point instead of the moment the diagnosis of osteoarthritis (OA) of the hip is established, therefore little is known about the length of time a patient diagnosed with hip OA stays under the care of a general practitioner (GP). No knowledge on factors of influence on this time period is available either. Aim of this study was thus to determine the time an incident hip OA patient stays in the care of a GP until referral to an orthopaedic department. Influencing factors were also analyzed. A prospective observational study was conducted based on data over a 10-year period from a general practice-based registration network (17 GPs, > 30,000 patients registered yearly). Patients with the diagnosis of hip OA were included. A survival analysis was used to determine time until referral to an orthopaedic department, and to determine factors of influence on this time. Of 391 patients diagnosed with hip OA, 121 (31%) were referred; average survival time until referral was 82.0 months (95% CI 76.6-87.5). Less contact with the GP for hip complaints before the diagnosis of hip OA was established resulted in a decreased time to referral. The results of this study show that patients with hip OA were under the care of a general practitioner, and thus in primary care, for a considerable amount of time once the diagnosis of hip OA was established.
    BMC Family Practice 06/2011; 12(1):48. DOI:10.1186/1471-2296-12-48 · 1.67 Impact Factor
  • Source
    • "The pool of practices fluctuates from year to year, but its composition is guarded to provide a representative sample of about 2% of the Dutch population regarding age and gender in comparison with Dutch National Statistics [17] and the practices are representative of all Dutch general practices with respect to geographical distribution and degree of urbanization [14,18,19]. For estimating morbidity rates, EMRs of individual GP-patient-contacts are grouped into episodes of care (concerning one health problem), as registered directly by the GPs (when available) or by constructing episodes with the validated application EPICON [20], grouping contacts with similar ICPC-codes less than two months apart. Individual patients' characteristics include age, sex, urban/rural residence (address density) and living in deprived areas (defined by GP-capitation fee). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Sexually transmitted infections (STI) care in the Netherlands is primarily provided by general practitioners (GPs) and specialized STI centers. STI surveillance is based on data from STI centers, which show increasing numbers of clients. Data from a GP morbidity surveillance network were used to investigate the distribution in the provision of STI care and the usefulness of GP data in surveillance. Data on STI-related episodes and STI diagnoses based on ICPC codes and, for chlamydia, prescriptions, were obtained from GP electronic medical records (EMRs) of the GP network and compared to data from STI centers from 2002 to 2007. Incidence rates were estimated for the total population in the Netherlands. The incidence of STI-consultations and -diagnoses increased substantially in recent years, both at GPs and STI centers. The increase in consultations was larger than the increase in diagnoses; Chlamydia incidence rose especially at STI centers. GPs were responsible for 70% of STI-related episodes and 80-85% of STI diagnoses. STI centers attract relatively younger and more often male STI-patients than GPs. Symptomatic STIs like Herpes genitalis and genital warts were more frequently diagnosed at GPs and chlamydia, gonorrhea and syphilis at STI centers. GPs fulfill an important role in STI care, complementary to STI centers. Case definitions of STI could be improved, particularly by including laboratory results in EMRs. The contribution of primary care is often overlooked in STI health care. Including estimates from GP EMRs can improve the surveillance of STIs.
    BMC Family Practice 05/2010; 11(1):39. DOI:10.1186/1471-2296-11-39 · 1.67 Impact Factor
Show more