Considerations on the feasibility of using retrospective database in the study of rare diseases / 国际药学研究杂志

Retrospective database is one of the data sources of real-world study, which has the advantages of abundant sample size, large amount of information and multi-dimension, and can support many types of quantitative re-search. On the one hand, rare diseases have the characteristics of low prevalence and the pathogenesis is unclear some-times. On the other hand, the lack of epidemiological investigation of rare diseases in China poses a challenge in the de-velopment and economic evaluation of rare disease drugs. This paper briefly describes the characteristics of the retrospec-tive database, and takes the health administrative data and electronic medical record as examples. Then, we discuss the feasibility of using such data to carry out epidemiological research and economic evaluation of rare diseases. Furthermore, we clarify some specific key points worth mentioning in advance in terms of study designs and protocols. Lastly, we pro-vide a reference for the subsequent use of retrospective data in conducting practical research.

High agreement between the new Mongolian electronic immunization register and written immunization records: a health centre based audit

Introduction: Monitoring of vaccination coverage is vital for the prevention and control of vaccine-preventable diseases. Electronic immunization registers have been increasingly adopted to assist with the monitoring of vaccine coverage; however, there is limited literature about the use of electronic registers in low- and middle-income countries such as Mongolia. We aimed to determine the accuracy and completeness of the newly introduced electronic immunization register for calculating vaccination coverage and determining vaccine effectiveness within two districts in Mongolia in comparison to written health provider records. Methods: We conducted a cross-sectional record review among children 2–23 months of age vaccinated at immunization clinics within the two districts. We linked data from written records with the electronic immunization register using the national identification number to determine the completeness and accuracy of the electronic register. Results: Both completeness (90.9%; 95% CI: 88.4–93.4) and accuracy (93.3%; 95% CI: 84.1–97.4) of the electronic immunization register were high when compared to written records. The increase in completeness over time indicated a delay in data entry. Conclusion: Through this audit, we have demonstrated concordance between a newly introduced electronic register and health provider records in a middle-income country setting. Based on this experience, we recommend that electronic registers be accompanied by routine quality assurance procedures for the monitoring of vaccination programmes in such settings.

A Preliminary Survey to Measure the Quality Indicators of End-of-life Cancer Care Using the Japanese National Database

Purpose: This survey aimed to develop a methodology for measuring the quality indicators of end-of-life care for cancer patients using the Japanese National Database, which was comprised of the health insurance claim data of all Japanese people. Methods: Life-sustaining treatment (LST) and chemotherapy near the time of death are accepted as reliable indicators of poor quality end-of-life care. To measure these, the Sampling Data Set (SDS) from the National Database (NDB) was used. Results: 1,233 cancer patients were studied, who had died from 14th to 31st October, 2012. The rates of LST and chemotherapy in the final 14 days of life were 8.2% (95%CI 6.7-10.1), 3.5% (2.6-4.8) for inpatients (n=1,079) respectively. In the SDS, 27-70% of chemotherapy drugs were not named, in order to prevent identification of patients receiving rare treatment. Discussion: The figures for rates of chemotherapy might be underestimated in the sampling data set, because of the anonymization of rare treatment. And in-patient and out-patient data may in some cases have been duplicated when entries applied to the same person. In the future using the NDB, it might be possible overcome some of these limitations.

Uma metodologia para explicar diferenças entre dados administrativos e pesquisas amostrais, com aplicação para o Bolsa Família e o Benefício de Prestação Continuada na PNAD / /A methodology for explaining differences between administrative data and findings from sample-based surveys regarding the Family Allowance Program and the Continuous Cash Benefit Program / Una metodología para explicar diferencias entre datos administrativos y encuestas por muestreo, con aplicación para el Bolsa Familia y el Beneficio de Prestación Continuada en la PNAD

Estimativas feitas com base em pesquisas domiciliares amostrais muitas vezes diferem bastante dos dados administrativos. Nas PNADs, o número estimado de beneficiários do Programa Bolsa Família (PBF) e do Benefício de Prestação Continuada (BPC) é sempre bem inferior ao número oficial. O objetivo deste artigo é apresentar uma metodologia simples, baseada nas características do desenho amostral das pesquisas domiciliares, para explicar essa diferença, decompondo-a em três termos: o viés de representatividade (derivado da escolha dos locais, áreas censitárias ou municípios para a pesquisa); o viés de captação (decorrente de problemas de captação nos locais selecionados); e a interação entre ambos. A aplicação dessa metodologia ao PBF e ao BPC mostra que, no primeiro caso, o viés de representatividade explica boa parte do problema: a seleção de municípios pesquisados é responsável por 40% da diferença observada entre os dados oficiais e os da PNAD. No caso do BPC, o viés de representatividade tenderia a agir no sentido oposto. Portanto, o viés de captação é inteiramente responsável pela diferença observada. Além disso, a declaração equivocada do BPC como benefício previdenciário na PNAD parece ocorrer, sobretudo, no período anterior a 2004 e não explica inteiramente o pequeno número de beneficiários identificados nas PNADs.

Estimates based on household surveys often differ considerably from administrative records. In the National Household Sample Surveys (PNADs) of the Brazilian Statistics Department, the estimated number of beneficiaries of the Family Allowance (Bolsa Família) Program (PBF), and of beneficiaries of the Continuous Cash Benefit (Benefício de Prestação Continuada) Program (BPC) is always lower than official figures. This paper presents a simple methodology, based on the sampling design of household surveys, to explain these differences, by decomposing them into three terms: the representativeness bias (derived from the choice of places, census tracts or municipalities for the survey); the data collection bias (derived from data collection problems at the chosen sites); and the interaction between them. The application of this methodology to the present cases shows that the representativeness bias of the PBF accounts for 40% of the difference between official records and the PNAD. For the BPC, the representativeness bias tends to act in the opposite direction. That is, In other words, the data collection bias is entirely responsible for the observed difference. Also, the erroneous reporting of the BPC as a Social Security benefit on the PNAD seems to have occurred mostly in the years prior to 2004, and does not entirely explain the low number of beneficiaries identified in the PNADs.

Estimaciones realizadas en base a encuestas de hogares por muestreo muchas veces difieren bastante de los datos administrativos. En la Pesquisa Nacional por Amostra de Domicílios - PNAD , el número estimado de beneficiarios del Programa Bolsa Familia (PBF) y del Beneficio de Prestación Continuada (BPC) siempre es bastante inferior al número oficial. El objetivo de este artículo es presentar una metodología simple, basada en las características del diseño por muestreo de las encuestas de hogares, para explicar esta diferencia, descomponiéndola en tres aspectos: el factor de representatividad (derivado de la elección de los sitios, áreas censitarias o municipios para la encuesta); el factor de captación (resultante de problemas de captación en los sitios seleccionados); y la interacción entre ambos. La aplicación de dicha metodología al PBF y al BPC muestra que, en el primer caso, el sesgo de representatividad explica una gran parte del problema: la selección de municipios investigados es la responsable por un 40% de la diferencia observada entre los datos oficiales y los de la PNAD. En el caso del BPC, el sesgo de representatividad tendería a actuar en el sentido opuesto. Por lo tanto, el sesgo de captación es totalmente responsable por la diferencia observada. Además, la declaración equivocada del BPC como beneficio de la previsión social en la PNAD parece ocurrir sobre todo en el periodo anterior a 2004 y no explica por completo el pequeño número de beneficiarios identificados en las PNADs.

Comparative Study on Three Algorithms of the ICD-10 Charlson Comorbidity Index with Myocardial Infarction Patients / 예방의학회지

OBJECTIVES: To compare the performance of three International Statistical Classification of Diseases, 10th Revision translations of the Charlson comorbidities when predicting in-hospital among patients with myocardial infarction (MI). METHODS: MI patients > or =20 years of age with the first admission during 2006 were identified(n=20,280). Charlson comorbidities were drawn from Heath Insurance Claims Data managed by Health Insurance Review and Assessment Service in Korea. Comparisions for various conditions included (a) three algorithms (Halfon, Sundararajan, and Quan algorithms), (b) lookback periods (1-, 3- and 5-years), (c) data range (admission data, admission and ambulatory data), and (d) diagnosis range (primary diagnosis and first secondary diagnoses, all diagnoses). The performance of each procedure was measured with the c-statistic derived from multiple logistic regression adjusted for age, sex, admission type and Charlson comorbidity index. A bootstrapping procedure was done to determine the approximate 95% confidence interval. RESULTS: Among the 20,280 patients, the mean age was 63.3 years, 67.8% were men and 7.1% died while hospitalized. The Quan and Sundararajan algorithms produced higher prevalences than the Halfon algorithm. The c-statistic of the Quan algorithm was slightly higher, but not significantly different, than that of other two algorithms under all conditions. There was no evidence that on longer lookback periods, additional data, and diagnoses improved the predictive ability. CONCLUSIONS: In health services study of MI patients using Health Insurance Claims Data, the present results suggest that the Quan Algorithm using a 1-year lookback involving primary diagnosis and the first secondary diagnosis is adequate in predicting in-hospital mortality.