Gumbel based p-value approximations for spatial scan statistics

Abstract

BackgroundThe spatial and space-time scan statistics are commonly applied for the detection of geographical disease clusters. Monte Carlo hypothesis testing is typically used to test whether the geographical clusters are statistically significant as there is no known way to calculate the null distribution analytically. In Monte Carlo hypothesis testing, simulated random data are generated multiple times under the null hypothesis, and the p-value is r/(R + 1), where R is the number of simulated random replicates of the data and r is the rank of the test statistic from the real data compared to the same test statistics calculated from each of the random data sets. A drawback to this powerful technique is that each additional digit of p-value precision requires ten times as many replicated datasets, and the additional processing can lead to excessive run times.ResultsWe propose a new method for obtaining more precise p-values with a given number of replicates. The collection of test statistics from the random replicates is used to estimate the true distribution of the test statistic under the null hypothesis by fitting a continuous distribution to these observations. The choice of distribution is critical, and for the spatial and space-time scan statistics, the extreme value Gumbel distribution performs very well while the gamma, normal and lognormal distributions perform poorly. From the fitted Gumbel distribution, we show that it is possible to estimate the analytical p-value with great precision even when the test statistic is far out in the tail beyond any of the test statistics observed in the simulated replicates. In addition, Gumbel-based rejection probabilities have smaller variability than Monte Carlo-based rejection probabilities, suggesting that the proposed approach may result in greater power than the true Monte Carlo hypothesis test for a given number of replicates.ConclusionsFor large data sets, it is often advantageous to replace computer intensive Monte Carlo hypothesis testing with this new method of fitting a Gumbel distribution to random data sets generated under the null, in order to reduce computation time and obtain much more precise p-values and slightly higher statistical power.

Spatial analysis of gastric cancer morbidity in regions of rapid urbanization: a case study in Xiamen, China

“…The maximum likelihood ratio window is the most likely spatial cluster of gastric cancer morbidity and Monte Carlo simulation was used to calculate the p value. The null hypothesis is that the gastric cancer morbidity is the same in all the sub-districts, compared with alternative hypothesis that the gastric cancer morbidity within the window is significantly higher than that outside (Kulldorff 1997;Abrams et al 2010). Sometimes it is interesting to see whether a significant cluster can decompose to several non-overlapping subclusters.…”

Spatial cluster detection for ordinal outcome data

“…The Gumbel‐based p ‐value is calculated as the probability that the test statistic is larger than the observed value, assuming the distribution of the test statistic under the null is Gumbel with the location and scale parameters estimated from MC replications. Abrams et al also showed that 99 replications for Gumbel approximation have about the same power with 999 replications for MC hypothesis testing. For comparison, we also list likelihood ratio test statistics calculated using the LRO‐based method for the clusters detected using the STO‐based method in Table .…”

Prospective space–time surveillance with cumulative surfaces for geographical identification of the emerging cluster

“…The reason for this is a theorem of Piterbarg (1996). The Gumbel distribution has also been used previously by Abrams et al (2010) to obtain an approximation to the spatial scan statistic. In the second approach, the threshold h emp is obtained as the percentile of the empirical distribution function of the maxima.…”