Koek Huberdina L, Kardaun Jan W P F, Gevers Evelien, de Bruin Agnes, Reitsma Joannes B, Grobbee Diederick E, Bots Michiel L
Julius Center for Health Sciences and Primary Care, University Medical Center, Utrecht, HPN Str. 06.131, PO Box 85.060, 3508 AB Utrecht, The Netherlands.
Eur J Epidemiol. 2007;22(11):755-62. doi: 10.1007/s10654-007-9174-6. Epub 2007 Sep 8.
To compare levels of and trends in incidence and hospital mortality of first acute myocardial infarction (AMI) based on routinely collected hospital morbidity data and on linked registers. Cases taken from routine hospital data are a mix of patients with recurrent and first events, and double counting occurs when cases are admitted for an event several times during 1 year. By linkage of registers, recurrent events and double counts can be excluded.
In 1995 and 2000, 28,733 and 25,864 admissions for AMI were registered in the Dutch national hospital discharge register. Linkage with the population register yielded 21,565 patients with a first AMI in 1995 and 20,414 in 2000.
In 1995 and 2000, the incidence based on the hospital register was higher than based on the linked registers in men (22% and 23% higher) and women (18% and 20% higher). In both years, hospital mortality based on the hospital register and on linked registers was similar. The decline in incidence between 1995 and 2000 was comparable whether based on standard hospital register data or linked data (18% and 20% in men, 15% and 17% in women). Similarly, the decline in hospital mortality was comparable using either approach (11% and 9% in both men and women).
Although the incidence based on routine hospital data overestimates the actual incidence of first AMI based on linked registers, hospital mortality and trends in incidence and hospital mortality are not changed by excluding recurrent events and double counts. Since trends in incidence and hospital mortality of AMI are often based on national routinely collected data, it is reassuring that our results indicate that findings from such studies are indeed valid and not biased because of recurrent events and double counts.
基于常规收集的医院发病数据和关联登记册，比较首次急性心肌梗死（AMI）的发病率及住院死亡率水平和趋势。从常规医院数据中获取的病例包含复发事件和首次事件的患者，当病例在1年内因同一事件多次入院时会出现重复计数。通过登记册的关联，可以排除复发事件和重复计数。
1995年和2000年，荷兰国家医院出院登记册中登记了28,733例和25,864例AMI入院病例。与人口登记册关联后，1995年有21,565例首次AMI患者，2000年有20,414例。
1995年和2000年，基于医院登记册的发病率高于基于关联登记册的发病率，男性分别高22%和23%，女性分别高18%和20%。在这两年中，基于医院登记册和关联登记册的住院死亡率相似。1995年至2000年发病率的下降幅度，基于标准医院登记册数据或关联数据时相当（男性分别为18%和20%，女性分别为15%和17%）。同样，使用任何一种方法，住院死亡率的下降幅度也相当（男性和女性均为11%和9%）。
尽管基于常规医院数据的发病率高估了基于关联登记册的首次AMI实际发病率，但排除复发事件和重复计数后，住院死亡率以及发病率和住院死亡率的趋势并未改变。由于AMI发病率和住院死亡率的趋势通常基于国家常规收集的数据，令人欣慰的是，我们的结果表明此类研究的结果确实有效，不会因复发事件和重复计数而产生偏差。
