METHODS: Studies were identified in the databases MEDLINE and Embase and screened for relevance before items were defined and extracted within the following categories: forecast methodology, measure of capacity, forecast horizon, healthcare setting, target diagnosis, validation methods, and implementation.
RESULTS: 84 studies were selected, all focusing on various capacity outcomes, including number of hospital beds/ patients, staffing, and length of stay. The selected studies employed different analytical models grouped in six items; discrete event simulation (N = 13, 15 %), generalized linear models (N = 21, 25 %), rate multiplication (N = 15, 18 %), compartmental models (N = 14, 17 %), time series analysis (N = 22, 26 %), and machine learning not otherwise categorizable (N = 12, 14 %). The review further provides insights into disease areas with infectious diseases (N = 24, 29 %) and cancer (N = 12, 14 %) being predominant, though several studies forecasted healthcare capacity needs in general (N = 24, 29 %). Only about half of the models were validated using either temporal validation (N = 39, 46 %), cross-validation (N = 2, 2 %) or/and geographical validation (N = 4, 5 %).
CONCLUSIONS: The forecasting models\' applicability can serve as a resource for healthcare stakeholders involved in designing future healthcare capacity estimation. The lack of routine performance validation of the used algorithms is concerning. There is very little information on implementation and follow-up validation of capacity planning models.
方法:在MEDLINE和Embase数据库中确定了研究,并在定义和提取以下类别的项目之前筛选了相关性:预测方法,衡量能力,预测范围,医疗保健设置,目标诊断,验证方法,和执行。
结果:选择了84项研究,所有这些都集中在各种能力成果上,包括医院病床/病人的数量,人员配备,和逗留时间的长短。选定的研究采用了分为六个项目的不同分析模型;离散事件模拟(N=13,15%),广义线性模型(N=21,25%),率倍增(N=15,18%),隔室模型(N=14,17%),时间序列分析(N=22,26%),和机器学习不可分类(N=12,14%)。该综述进一步提供了以传染病(N=24,29%)和癌症(N=12,14%)为主的疾病领域的见解,尽管有几项研究预测了总体上的医疗保健能力需求(N=24,29%)。只有大约一半的模型使用任一时间验证进行了验证(N=39,46%),交叉验证(N=2,2%)或/和地理验证(N=4,5%)。
结论:预测模型的适用性可以作为参与设计未来医疗保健能力估计的医疗保健利益相关者的资源。所使用的算法缺乏常规性能验证是令人担忧的。关于容量规划模型的实施和后续验证的信息很少。