Understanding patient choice is vital in assessing the closeness of competition between hospitals. The standard technique used in the UK is to estimate substitution patterns based on historical general practitioner (GP) referrals. In this paper we compare the results of the ‘GP referral’ methodology to a demand estimation approach. Using patient-level data over a 3-year period (2012/13–2014/15) we apply both methodologies to every hypothetical merger between hospitals in England, for three specialties. We find a high degree of consistency between the two approaches, suggesting that GP referral analysis is a useful and reliable filter in merger cases. There are a small number of cases, however, in which the GP referral approach filters out potentially problematic mergers. This is particularly true for ‘borderline’ cases. Filtering should therefore be done with caution and in conjunction with additional evidence.

中文翻译：

---

估计医院合并中的分流比率

了解患者的选择对于评估医院之间竞争的紧密程度至关重要。英国使用的标准技术是根据历史全科医生（GP）推荐来估算替代模式。在本文中，我们将“ GP转诊”方法的结果与需求估算方法进行了比较。我们使用3年期（2012 / 13–2014 / 15）的患者水平数据，将这两种方法应用于英格兰三个医院之间每个假设的三个专科合并。我们发现这两种方法之间具有高度的一致性，这表明在合并案例中，GP推荐分析是一种有用且可靠的过滤器。但是，在少数情况下，GP推荐方法可以过滤出可能存在问题的合并。对于“边界”案件尤其如此。