A power comparison of generalized additive models and the spatial scan statistic in a casecontrol setting
 Robin L Young^{1}Email author,
Affiliated with
 Janice Weinberg^{1},
Affiliated with
 Verónica Vieira^{2},
Affiliated with
 Al Ozonoff^{1} and
Affiliated with
 Thomas F Webster^{2}
Affiliated with
DOI: 10.1186/1476072X937
© Young et al. 2010
Received: 3 February 2010
Accepted: 19 July 2010
Published: 19 July 2010
Abstract
Background
A common, important problem in spatial epidemiology is measuring and identifying variation in disease risk across a study region. In application of statistical methods, the problem has two parts. First, spatial variation in risk must be detected across the study region and, second, areas of increased or decreased risk must be correctly identified. The location of such areas may give clues to environmental sources of exposure and disease etiology. One statistical method applicable in spatial epidemiologic settings is a generalized additive model (GAM) which can be applied with a bivariate LOESS smoother to account for geographic location as a possible predictor of disease status. A natural hypothesis when applying this method is whether residential location of subjects is associated with the outcome, i.e. is the smoothing term necessary? Permutation tests are a reasonable hypothesis testing method and provide adequate power under a simple alternative hypothesis. These tests have yet to be compared to other spatial statistics.
Results
This research uses simulated point data generated under three alternative hypotheses to evaluate the properties of the permutation methods and compare them to the popular spatial scan statistic in a casecontrol setting. Case 1 was a single circular cluster centered in a circular study region. The spatial scan statistic had the highest power though the GAM method estimates did not fall far behind. Case 2 was a single point source located at the center of a circular cluster and Case 3 was a line source at the center of the horizontal axis of a square study region. Each had linearly decreasing logodds with distance from the point. The GAM methods outperformed the scan statistic in Cases 2 and 3. Comparing sensitivity, measured as the proportion of the exposure source correctly identified as high or low risk, the GAM methods outperformed the scan statistic in all three Cases.
Conclusions
The GAM permutation testing methods provide a regressionbased alternative to the spatial scan statistic. Across all hypotheses examined in this research, the GAM methods had competing or greater power estimates and sensitivities exceeding that of the spatial scan statistic.
Background
Statistical tests applied in spatial epidemiology have two primary purposes. The first is to detect spatial variation in disease risk across a study region and the second is to identify areas of increased or decreased risk [1–4]. We consider generalized additive models (GAMs) and the spatial scan statistic; two popular methods that can be used for both purposes.
GAMs are generalizations of generalized linear models that allow nonparametric functions of covariates to be modeled in an additive framework [5]. Webster et al. (2006) used GAMs in spatial settings with a bivariate locally weighted regression (LOESS) smooth [5] and performed hypothesis tests using permutation techniques to determine whether there was spatial variation in disease risk and to locate statistically significant areas of increased or decreased risk [6]. Similar methods have been applied by other authors using tests based on permutation, bootstrap, and Monte Carlo techniques [7–10].
In previous research, we evaluated four permutation tests applied with GAMs to determine type I error rates and power estimates under simple hypotheses (Young, Weinberg, Vieira, Ozonoff, Webster: The Power of Hypothesis Testing Using Generalized Additive Models with Bivariate Smoothers, submitted) [11]. The four methods differed primarily in the determination of the span (neighborhood) size when applying GAMs to observed and permuted datasets. For the conditional permutation test (CPT), originally proposed by Webster et al. (2006), we selected an optimal span by applying GAMs to observed data using a range of possible span sizes. Akaike's Information Criterion (AIC) was recorded for each model and the minimal model AIC corresponded to the optimal span [6, 12]. The statistic of interest, the difference in deviances of models including and excluding the bivariate LOESS smoothing term, was recorded for the observed data. GAMs were then applied to permuted datasets using the optimal span selected for the observed data to produce a conditional permutation distribution of difference in deviance statistics. We determined significance through the comparison of the observed statistic to the sampling distribution generated from the repeated permutations [6]. This method had an inflated type I error rate when applied using the nominal α cutoff. For a nominal significance level of 0.05, CPT had an estimated type I error rate of 9.5% when applied with a bivariate smoothing term. When the null hypothesis was rejected for pvalues less than 0.025, the observed type I error rate fell within a 95% confidence interval of 0.05, the desired significance level [11].
Description of Hypothesis Testing Methods and Significance Cutoffs
Hypothesis Testing Method  Abbreviation  Description  Significance Cutoff 

Conditional Permutation Test  CPT  Select optimal span size for observed data by minimizing AIC statistic across range of spans. Compare difference in deviance statistic to conditional permutation distribution obtained by holding span size constant.  0.025 
Fixed Span Permutation Test  FSPT  Select span size a priori. Compare difference in deviance statistic to conditional permutation distribution obtained by holding span size constant.  0.05 
Fixed Multiple Span Permutation Test  FMSPT  Select 35 span sizes a priori. For each span size, compare the difference in deviance statistic to corresponding conditional permutation distribution obtained by holding the span size constant. Reject the null hypothesis if at least one pvalue falls below the significance cutoff. 

Unconditional Permutation Test  UPT  Select optimal span size for observed data as in CPT. Compare difference in deviance statistic to unconditional permutation distribution obtained by selecting optimal span size for each permuted dataset.  0.05 
Spatial Scan Statistic    Detects the most likely cluster through a likelihood ratio test comparing the likelihood of cases within to outside a circular zone of interest. Pvalues are obtained through Monte Carlo methods  0.05 
The spatial scan statistic, a popular method proposed by Kulldorff and Nargawalla (1995), detects the most likely cluster through comparison of likelihoods of cases falling within and outside circular zones [13]. In recent power evaluations, the scan statistic performed well with a single circular cluster [2, 14, 15] but underperformed with multiple and noncircular clusters [16]. When applied to casecontrol data, aside from stratified analyses, the scan statistic cannot be adjusted for covariates [16]. We applied the scan statistic through SaTScan, publicly available free software [17], to reflect its application in spatial statistics and spatial epidemiology.
GAMs and the scan statistic were compared in a previous study that focused on cluster detection using aggregate data. The performance of the two methods depended greatly on the shape of the cluster and with irregularly shaped clusters, GAMs outperformed the scan statistic [18]. In this study, we applied the CPT, FMSPT, and spatial scan statistic to simulated casecontrol point data to estimate power for global and sensitivity for local hypothesis tests. The CPT and FMSPT were selected for comparison as they are computationally efficient, had similar power estimates in a previous study (Young, Weinberg, Vieira, Ozonoff, Webster: The Power of Hypothesis Testing Using Generalized Additive Models with Bivariate Smoothers, submitted), and neither method requires a priori selection of a single span.
Methods
Simulated Data
Simulated data had a dichotomous outcome and geographic locations generated from a bivariate uniform distribution of longitude and latitude. Odds ratios were chosen to produce a wide range of theoretical power. Odds ratios less than 1.0 indicate areas of decreased risk while odds ratios greater than 1.0 indicate areas of increased risk. For each set of parameters, 1000 datasets were simulated, each containing 1000 observations, chosen to reflect previous studies performed on the Cape Cod Family Health Study data that used GAMs as a statistical analysis technique [19–21]. Statistical analyses were applied to point data and the nominal α level was 0.05.
Simulated data for Case 2 reflected a circular study region with linearly increasing or decreasing logodds of disease with proximity to the center of the region. (Figure 2) This was a simplified pattern of what may be observed where increased proximity to some point, such as a lead smelter [22], increased the risk of disease. The probability of disease at the edge of the region, i.e. the subjects least exposed, was equal to 5 or 20% and odds ratios comparing subjects at the center to those at the edge of the region included 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, and 3.5.
In Case 3, the study region was square with logodds of disease increasing or decreasing linearly with proximity to the center of the horizontal axis of the study region. (Figure 3) These data followed a simple pattern similar to increased risk of disease with proximity to heavytraffic roadways [23, 24]. As with Case 2, the probabilities of disease for those least exposed, i.e. living at the horizontal edges of the region, were equal to either 5 or 20%. There was no variation in disease risk across the vertical axis. The odds ratios were 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, and 3.5 comparing subjects living at the center to those at the edge of the horizontal axis.
Theoretical Power
Data for Case 1 could be appropriately analyzed using a Pearson chisquare test while data for Cases 2 and 3 could be appropriately analyzed using simple logistic regressions. We derived the theoretical power for each Case to determine how the spatial methods compare to the more simple tests. Details of these derivations are available in Additional file 1. 95% confidence intervals were computed for each power estimate. The margin of error was computed using the standard deviation of the estimated power, i.e. , where is the estimated power.
Generalized Additive Models (GAMs)
We applied GAMs to simulated data using a bivariate LOESS smoothing term to adjust for geographic location [6] using the gam package [25] available in R v2.8.0 [26]. Two hypothesis tests were performed using the GAM framework: the CPT and FMSPT. When performing the CPT, an optimal span size for the observed data was selected through the application of GAM models to the observed data using a range of span sizes between 0.05 and 0.95. The AIC was recorded for each model and the span was selected to minimize the model AIC [6, 12]. The span size was held constant as GAMs were applied to 999 permuted datasets. The statistic of interest was the difference in deviances between models including and excluding the smoothed term for geographic location. The null hypothesis was rejected if the observed difference in deviance statistic fell in the upper 2.5% of the distribution of statistics from permuted datasets [6, 11].
The FMSPT was performed through the application of GAM models across either three or five predetermined span sizes, denoted by FMSPT3 and FMSPT5 respectively. The three span sizes selected were 0.1, 0.5, and 0.9 while the five span sizes were 0.1, 0.3, 0.5, 0.7, and 0.9. From each model the difference in deviance statistic was recorded and subsequently compared to the permutation distribution with the same span size. For FMSPT3, the null hypothesis was rejected if the difference in deviance statistic fell in the upper of the distribution while for FMSPT5, it was rejected if the statistic fell in the upper of the distribution [27].
Syntax used to generate data for this study and to minimize the AIC statistic across multiple span sizes using the software R v2.8.0 [26] is available on the Boston University Superfund Research Program website at: http://www.busbrp.org/projects/project2.html.
Spatial Scan Statistic
The spatial scan statistic is a method that overlays the study region with overlapping circular regions centered at observation locations with radii varying continuously from zero to some specified upper limit (here, radii vary from zero to containing onehalf of the study population). A zone is the infinite number of circles centered at some arbitrary point location with radii varying from zero to the upper bound. For a given radius, a zone can be further described by the number of individuals and cases falling inside the circle. The spatial scan statistic tests the null hypothesis of spatial randomness, i.e. the probability of disease within a circular zone equals that outside the zone, through a likelihood ratio test and detects the most likely cluster as the zone that maximizes the likelihood under the full parameter space. The distribution of the likelihood ratio depends on the underlying population distribution, upon which no assumptions have been made. With small samples it is possible to find the exact distribution; however for larger datasets Monte Carlo simulations are required [13].
In this study, the scan statistic was applied through the free available software SaTScan v 7.0.3 [17] using a purely spatial Bernoulli model, appropriate for casecontrol data. We rejected the null hypothesis if the most likely cluster was significant at the 0.05 level.
Detecting Exposure Source Locations
We aimed to evaluate the ability of the GAM permutation tests and the spatial scan statistic to correctly identify the exposure source as high or low risk, i.e. the sensitivity of the methods. For Cases 1 and 2, we defined the CPT as successful in locating the exposure source if the global null hypothesis was rejected and the pointwise predicted logodds for the center of the study region fell in either the upper or lower 2.5% of the pointwise permutation distribution of predicted logodds. We considered the FMSPT successful if the global null hypothesis was rejected and at least one test predicted a pointwise logodds falling in the upper or lower 2.5% of the corresponding predicted permutation distribution. For Case 1, we also examined the proportion of the true cluster correctly identified as high or low risk by the pointwise tests, given that the global hypothesis was rejected. For Case 3, we defined sensitivity as the proportion of the vertical exposure source that was detected as increased risk for datasets when the global null hypothesis was rejected. For the FMSPTs, we present the proportion of the vertical source detected by at least one span size.
For Cases 1 and 2, if the center of the region was detected as part of a significant most likely cluster at the 0.05 level, the scan statistic was considered successful in identifying the area of risk. For Case 1, we also examined the percent of the true cluster overlapped by a significant most likely cluster. For Case 3, we examined the proportion of the vertical exposure source included in a significant most likely cluster.
The sensitivity to detect the exposure source is undefined for data simulated under the null hypothesis and so sensitivity estimates for these data are excluded. The local hypothesis tests are exploratory in nature but provide an additional measure by which the tests can be compared. We did not examine a measure of specificity in this analysis as, for Cases 2 and 3, the exposure source was not dichotomous in nature.
Results
Theoretical Power Based on Pearson ChiSquare Test and Simple Logistic Regression
Probability of Disease Unexposed  Odds Ratios  

0.5  1.0  1.5  2.0  2.5  3.0  3.5  
Case 1^{^}  0.05  0.258  0.050  0.214  0.598  0.884  >0.999  
0.20  0.731  0.050  0.521  0.953  0.999  >0.999  
Case 2*  0.05  0.255  0.050  0.167  0.331  0.502  0.650  0.766 
0.20  0.565  0.050  0.321  0.663  0.872  0.959  0.988  
Case 3*  0.05  0.311  0.050  0.215  0.464  0.695  0.850  0.935 
0.20  0.691  0.050  0.432  0.832  0.970  0.996  >0.999 
Observed Power for GAM Hypothesis Tests and Spatial Scan Statistic
Probability of Disease Unexposed  Odds Ratios  

0.5  1.0  1.5  2.0  2.5  3.0  3.5  
Case 1  Power (95% CI)  Power (95% CI)  Power (95% CI)  Power (95% CI)  Power (95% CI)  Power (95% CI)  Power (95% CI)  
0.05  CPT  0.081 (0.0640.098)  0.043 (0.0300.056)  0.070 (0.0540.086)  0.174 (0.1510.197)  0.301 (0.2730.329)  0.474 (0.4430.505)  
FMSPT3  0.055 (0.0410.069)  0.033 (0.0220.044)  0.048 (0.0350.061)  0.139 (0.1180.160)  0.250 (0.2230.277)  0.403 (0.3730.433)  
FMSPT5  0.044 (0.0310.057)  0.023 (0.0140.032)  0.039 (0.0270.051)  0.103 (0.0840.122)  0.225 (0.1990.251)  0.361 (0.3310.391)  
Scan Statistic  0.087 (0.0700.104)  0.052 (0.0380.066)  0.075 (0.0590.091)  0.169 (0.1460.192)  0.302 (0.2740.330)  0.494 (0.4630.525)  
0.20  CPT  0.239 (0.2130.265)  0.047 (0.0340.06)  0.149 (0.1270.171)  0.447 (0.4160.478)  0.764 (0.7380.79)  0.923 (0.9060.94)  
FMSPT3  0.166 (0.1430.189)  0.027 (0.0170.037)  0.111 (0.0920.130)  0.378 (0.3480.408)  0.697 (0.6690.725)  0.890 (0.8710.909)  
FMSPT5  0.152 (0.1300.174)  0.020 (0.0110.029)  0.094 (0.0760.112)  0.33 (0.3010.359)  0.673 (0.6440.702)  0.880 (0.8600.900)  
Scan Statistic  0.243 (0.2160.270)  0.044 (0.0310.057)  0.121 (0.1010.141)  0.467 (0.4360.498)  0.833 (0.8100.856)  0.963 (0.9510.975) 
Case 2  

0.05  CPT  0.095 (0.0770.113)  0.069 (0.0530.085)  0.054 (0.0400.068)  0.094 (0.0760.112)  0.176 (0.1520.200)  0.230 (0.2040.256)  0.381 (0.3510.411) 
FMSPT3  0.074 (0.0580.090)  0.048 (0.0350.061)  0.035 (0.0240.046)  0.075 (0.0590.091)  0.131 (0.1100.152)  0.186 (0.1620.210)  0.31 (0.2810.339)  
FMSPT5  0.053 (0.0390.067)  0.037 (0.0250.049)  0.029 (0.0190.039)  0.061 (0.0460.076)  0.115 (0.0950.135)  0.154 (0.1320.176)  0.281 (0.2530.309)  
Scan Statistic  0.059 (0.0440.074)  0.052 (0.0380.066)  0.048 (0.0350.061)  0.091 (0.0730.109)  0.137 (0.1160.158)  0.194 (0.1690.219)  0.293 (0.2650.321)  
0.20  CPT  0.223 (0.1970.249)  0.05 (0.0360.064)  0.118 (0.0980.138)  0.293 (0.2650.321)  0.520 (0.4890.551)  0.714 (0.6860.742)  0.878 (0.8580.898) 
FMSPT3  0.173 (0.1500.196)  0.033 (0.0220.044)  0.089 (0.0710.107)  0.22 (0.1940.246)  0.441 (0.4100.472)  0.658 (0.6290.687)  0.833 (0.8100.856)  
FMSPT5  0.147 (0.1250.169)  0.023 (0.0140.032)  0.073 (0.0570.089)  0.199 (0.1740.224)  0.404 (0.3740.434)  0.611 (0.5810.641)  0.798 (0.7730.823)  
Scan Statistic  0.143 (0.1210.165)  0.045 (0.0320.058)  0.080 (0.0630.097)  0.17 (0.1470.193)  0.367 (0.3370.397)  0.584 (0.5530.615)  0.758 (0.7310.785) 
Case 3  

0.05  CPT  0.094 (0.0760.112)  0.040 (0.0280.052)  0.059 (0.0440.074)  0.131 (0.1100.152)  0.240 (0.2140.266)  0.426 (0.3950.457)  0.547 (0.5160.578) 
FMSPT3  0.071 (0.0550.087)  0.027 (0.0170.037)  0.037 (0.0250.049)  0.078 (0.0610.095)  0.178 (0.1540.202)  0.359 (0.3290.389)  0.478 (0.4470.509)  
FMSPT5  0.056 (0.0420.070)  0.017 (0.0090.025)  0.028 (0.0180.038)  0.064 (0.0490.079)  0.152 (0.1300.174)  0.324 (0.2950.353)  0.437 (0.4060.468)  
Scan Statistic  0.068 (0.0520.084)  0.042 (0.0300.054)  0.059 (0.0440.074)  0.085 (0.0680.102)  0.129 (0.1080.15)  0.201 (0.1760.226)  0.285 (0.2570.313)  
0.20  CPT  0.276 (0.2480.304)  0.058 (0.0440.072)  0.137 (0.1160.158)  0.400 (0.3700.430)  0.712 (0.6840.740)  0.882 (0.8620.902)  0.970 (0.9590.981) 
FMSPT3  0.218 (0.1920.244)  0.036 (0.0240.048)  0.094 (0.0760.112)  0.327 (0.2980.356)  0.622 (0.5920.652)  0.843 (0.8200.866)  0.954 (0.9410.967)  
FMSPT5  0.191 (0.1670.215)  0.026 (0.0160.036)  0.074 (0.0580.090)  0.298 (0.2700.326)  0.587 (0.5560.618)  0.818 (0.7940.842)  0.942 (0.9280.956)  
Scan Statistic  0.158 (0.1350.181)  0.047 (0.0340.060)  0.067 (0.0520.082)  0.192 (0.1680.216)  0.348 (0.3180.378)  0.534 (0.5030.565)  0.703 (0.6750.731) 
In general, FMSPT3 had higher power estimates than FMSPT5, perhaps due to the slightly conservative Bonferronilike significance cutoff adjustments. The greatest power estimates for the FMSPTs with an odds ratio of 3.0 were observed in Case 1 with a probability of disease for unexposed subjects of 0.20 and power estimates ranging from 0.027 (95%CI: 0.0170.037) to 0.890 (95%CI: 0.8710.909) and 0.020 (95%CI: 0.0110.029) to 0.880 (95%CI: 0.8600.900) for the 3 and 5 span tests, respectively. The power estimates for Case 3 were greater than those of Case 2 with maximal power estimates of 0.954 (95%CI: 0.9410.967) and 0.942 (95%CI: 0.9280.956) observed for FMSPT3 and FMSPT5, respectively, for an odds ratio of 3.5 in Case 3 and 0.833  (95%CI: 0.8100.856) and 0.798 (95%CI: 0.7730.823) in Case 2. (Table 3)
The spatial scan statistic performed best in Case 1, followed by Cases 2 and 3. In Case 1, the scan statistic had a maximum estimated power of 0.963 (95%CI: 0.9510.975) with a probability of disease outside the cluster of 0.20. In Cases 2 and 3 the maximum power was 0.758 (95%CI: 0.7310.785) and 0.703 (95%CI: 0.6750.731), respectively, for an odds ratio of 3.5. The test had an appropriate type I error rate for all cases and scenarios. It was outperformed by all three of the permutation testing methods in Cases 2 and 3 but had the highest power estimates in Case 1. (Table 3)
Case 1 Sensitivity  Mean Proportion of True Cluster Detected as Hot or Coldspot
Odds Ratios  

0.5  1.5  2.0  2.5  3.0  
Probability of Disease Unexposed  Mean (SD)  Mean (SD)  Mean (SD)  Mean (SD)  Mean (SD)  
0.05  CPT  0.380 (0.287)  0.360 (0.240)  0.582 (0.258)  0.671 (0.221)  0.740 (0.202) 
FMSPT3  0.418 (0.493)  0.513 (0.256)  0.697 (0.259)  0.807 (0.179)  0.855 (0.150)  
FMSPT5  0.364 (0.481)  0.538 (0.249)  0.725 (0.261)  0.834 (0.169)  0.876 (0.133)  
Scan Statistic  0.335 (0.435)  0.325 (0.331)  0.536 (0.359)  0.607 (0.320)  0.686 (0.290)  
0.20  CPT  0.653 (0.251)  0.516 (0.278)  0.738 (0.218)  0.811 (0.172)  0.855 (0.140) 
FMSPT3  0.584 (0.493)  0.644 (0.288)  0.851 (0.156)  0.926 (0.104)  0.961 (0.063)  
FMSPT5  0.618 (0.486)  0.660 (0.292)  0.874 (0.141)  0.935 (0.098)  0.969 (0.058)  
Scan Statistic  0.684 (0.333)  0.498 (0.380)  0.688 (0.303)  0.782 (0.242)  0.842 (0.186) 
Cases 2 and 3 Sensitivity  Detecting the Exposure Source Location
Probability of Disease Unexposed  Odds Ratios  

0.5  1.5  2.0  2.5  3.0  3.5  
Case 2 ^{ ^ }  Sensitivity (95% CI)  Sensitivity (95% CI)  Sensitivity (95% CI)  Sensitivity (95% CI)  Sensitivity (95% CI)  Sensitivity (95% CI)  
0.05  CPT  0.453 (0.4220.484)  0.352 (0.3220.382)  0.543 (0.5120.574)  0.688 (0.6590.717)  0.896 (0.8770.915)  0.882 (0.8620.902) 
FMSPT3  0.541 (0.5100.572)  0.543 (0.5120.574)  0.773 (0.7470.799)  0.824 (0.8000.848)  0.952 (0.9390.965)  0.952 (0.9390.965)  
FMSPT5  0.585 (0.5540.616)  0.517 (0.4860.548)  0.754 (0.7270.781)  0.8 (0.7750.825)  0.968 (0.9570.979)  0.947 (0.9330.961)  
Scan Statistic  0.026 (0.0160.036)  0.012 (0.0050.019)  0.038 (0.0260.050)  0.094 (0.0760.112)  0.153 (0.1310.175)  0.229 (0.2030.255)  
0.20  CPT  0.740 (0.7130.767)  0.559 (0.5280.590)  0.860 (0.8380.882)  0.950 (0.9360.964)  0.976 (0.9670.985)  0.987 (0.9800.994) 
FMSPT3  0.936 (0.9210.951)  0.652 (0.6220.682)  0.964 (0.9520.976)  0.995 (0.9910.999)  0.995 (0.9910.999)  0.998 (0.9951.000)  
FMSPT5  0.973 (0.9630.983)  0.644 (0.6140.674)  0.965 (0.9540.976)  0.995 (0.9910.999)  0.997 (0.9941.000)  0.999 (0.9971.000)  
Scan Statistic  0.082 (0.0650.099)  0.029 (0.0190.039)  0.12 (0.1000.140)  0.301 (0.2730.329)  0.533 (0.5020.564)  0.725 (0.6970.753)  
Case 3*  Mean (SD)  Mean (SD)  Mean (SD)  Mean (SD)  Mean (SD)  Mean (SD)  
0.05  CPT  0.256 (0.097)  0.234 (0.069)  0.333 (0.134)  0.391 (0.190)  0.463 (0.258)  0.524 (0.293) 
FMSPT3  0.308 (0.101)  0.344 (0.072)  0.436 (0.129)  0.516 (0.213)  0.59 (0.301)  0.634 (0.334)  
FMSPT5  0.368 (0.101)  0.353 (0.065)  0.473 (0.126)  0.549 (0.209)  0.622 (0.306)  0.661 (0.343)  
Scan Statistic  0.263 (0.319)  0.189 (0.234)  0.232 (0.261)  0.299 (0.279)  0.315 (0.268)  0.346 (0.277)  
0.20  CPT  0.394 (0.203)  0.339 (0.134)  0.454 (0.253)  0.548 (0.286)  0.641 (0.255)  0.704 (0.192) 
FMSPT3  0.511 (0.228)  0.453 (0.145)  0.574 (0.287)  0.659 (0.340)  0.737 (0.296)  0.788 (0.204)  
FMSPT5  0.550 (0.229)  0.475 (0.136)  0.598 (0.290)  0.638 (0.355)  0.758 (0.317)  0.807 (0.222)  
Scan Statistic  0.363 (0.280)  0.242 (0.280)  0.282 (0.300)  0.390 (0.287)  0.441 (0.267)  0.443 (0.257) 
In Case 3, the FMSPTs outperformed the CPT and scan statistic detecting a higher proportion of the vertical exposure source. For probabilities of disease of 0.05 and 0.20, FMSPT3 detected an average proportion of 0.634 (SD: 0.334) and 0.788 (SD: 0.204) and FMSPT5 detected 0.661 (SD: 0.343) and 0.807 (SD: 0.222) of the vertical exposure source, respectively, for an odds ratio of 3.5. The CPT detected proportions of 0.524 (SD: 0.293) and 0.704 (SD: 0.192) on average while the scan statistic had mean detection proportions of 0.346 (SD: 0.277) and 0.443 (SD: 0.257) for the probabilities of disease 0.05 and 0.20, respectively. (Table 5)
Radii of Scan Statistic Most Likely Cluster with Significant PValue (p < 0.05)
Odds Ratios  

0.5  1.0  1.5  2.0  2.5  3.0  3.5  
Mean (SD;MinMax)  Mean (SD;MinMax)  Mean (SD;MinMax)  Mean (SD;MinMax)  Mean (SD;MinMax)  Mean (SD;MinMax)  Mean (SD;MinMax)  
Case 1  
0.05  0.39 (0.25;0.041.12)  0.34 (0.26;0.040.95)  0.32 (0.24;0.031.09)  0.37 (0.19;0.030.91)  0.35 (0.16;0.031.02)  0.37 (0.14;0.020.91)  
0.20  0.37 (0.14;0.050.88)  0.24 (0.13;0.080.68)  0.33 (0.19;0.040.95)  0.37 (0.13;0.061.03)  0.38 (0.09;0.070.71)  0.38 (0.07;0.070.80)  
Case 2  
0.05  0.37 (0.23;0.040.70)  0.33 (0.24;0.050.96)  0.31 (0.22;0.041.01)  0.36 (0.26;0.031.10)  0.43 (0.24;0.041.09)  0.43 (0.2;0.040.84)  0.45 (0.19;0.041.00) 
0.20  0.4 (0.21;0.041.12)  0.28 (0.18;0.070.82)  0.36 (0.21;0.070.77)  0.45 (0.19;0.070.84)  0.47 (0.18;0.060.71)  0.50 (0.16;0.070.92)  0.53 (0.14;0.060.73) 
Case 3  
0.05  0.4 (0.29;0.040.95)  0.43 (0.32;0.061.06)  0.34 (0.26;0.051.03)  0.35 (0.24;0.060.94)  0.41 (0.24;0.040.85)  0.47 (0.22;0.051.02)  0.46 (0.22;0.041.10) 
0.20  0.46 (0.22;0.061.1)  0.38 (0.22;0.090.88)  0.39 (0.24;0.081.36)  0.41 (0.24;0.061.13)  0.49 (0.22;0.061.05)  0.52 (0.20;0.070.89)  0.52 (0.19;0.070.92) 
Discussion
Simulated data were used to compare the power and sensitivity of the CPT and FMSPTs performed with GAMs to the spatial scan statistic under three simple alternative hypotheses. Theoretical power was computed for each alternative hypothesis to provide a comparison of spatial statistic hypothesis tests to simpler methods.
In Case 1, a circular cluster was centered in the study region. The spatial scan statistic identifies clusters by placing circular zones across the region of interest and comparing the likelihood of disease within to outside the zones. As this method is similar to the pattern of disease risk for this Case, it is unsurprising that the scan statistic had the highest estimated power, nearing the theoretical power calculated for a Pearson chisquare test. The CPT had slightly lower power than the scan statistic and FMSPT3 and FMSPT5 had lower power estimates.
In Case 2, there was a linear association between Euclidean distance from the center of the circular study region and the logodds of disease. Case 3 was a square study region with a linear association between the proximity to the center of the horizontal axis and the logodds of disease. As these cases would be appropriately analyzed by logistic regression methods, GAM permutation tests had an advantage over the scan statistic in its flexibility to detect different patterns in disease risk. In both Cases 2 and 3, the CPT had the highest estimated power though estimates were at least 10% smaller than the theoretical power of a logistic regression. The scan statistic had the lowest power for Cases 2 and 3. For all tests, power estimates for Case 1 exceeded those of Cases 2 and 3. For the GAM permutation tests, power estimates for Case 3 were greater than those of Case 2 under similar conditions while the estimates were comparable between the two cases for the scan statistic.
The size of the most likely cluster and hot and coldspots identified by the scan statistic and GAM methods varied greatly across datasets, as observed in Figures 6 and 7. For Case 1, an odds ratio of 3.0, and a lower prevalence, i.e. a probability of disease outside the cluster of 0.05, the spatial scan statistic had larger variation in most likely cluster radius and a greater probability of a most likely cluster having a radius smaller than the true cluster than for higher prevalence. (Table 6, Figure 5a) The scan statistic showed a tendency to detect small clusters while the CPT tended to smooth over small variations in disease risk as a large span (span > 0.80) was more likely to be selected when analyzing diseases of lower prevalence. (Figure 5b)
Comparing model sensitivities, in Case 1, the FMSPT5 consistently detected the highest proportion of the true cluster as a hot or coldspot, followed by FMSPT3 and CPT with the scan statistic having the lowest mean proportion detected. It is not surprising that FMSPT3 and FMSPT5 had the highest sensitivity estimates as the definition of sensitivity of these tests considered points detected if they were considered a hot or coldspot in at least one of 3 or 5 models. Sensitivity for the CPT required the points to be detected at a single span size.
Of interest, the spatial scan statistic had the highest power estimates for Case 1 though it did not detect the highest proportion of the true cluster. As for its sensitivity, the scan statistic detected a most likely cluster of the correct size with a radius within ±0.01 of the true cluster radius in 19.4% of datasets with an odds ratio of 3.0 and a probability of disease outside the cluster of 0.20. Of these most likely clusters, 12.4% were centered in the correct location and only one dataset was observed to have a correct cluster radius and location with a pvalue of less than 0.05.
In Case 2, sensitivity was measured by the probability of detecting the exposure source, given that the global null hypothesis was rejected. In practice, after detecting variation in disease risk, public health resources may be sent to specific locations detected as hot or coldspots to determine the source of exposure. If the exposure point source is not included in the most likely cluster or hot/coldspot detected, it is unlikely that public health officials will be able identify the true exposure that is increasing disease risk. A minimum sensitivity of 80% may be considered a reasonable requirement of tests used for application. The FMSPTs had sensitivity estimates of at least 80% for odds ratios over 2.0 while the CPT had sensitivities of 80% for odds ratios of at least 3.0 for both probabilities of disease. The sensitivity of the scan statistic did not reach 80% for any odds ratios, having much lower estimates than the permutation testing methods. Of the datasets where the scan statistic detected a most likely cluster with a pvalue of less than 0.05, it rarely identified the correct exposure point source.
Sensitivity for Case 3 was measured as the proportion of the vertical exposure source identified as high or low risk, given that the global null hypothesis was rejected. Again, the spatial scan statistic had much lower sensitivity than the permutation testing methods. For odds ratios of at least 3.0 and a probability of disease for unexposed subjects of 0.20, the FMSPTs had sensitivity estimates of at least 70%, slightly lower than the desired magnitude. FMSPT5 had the highest sensitivity, followed by FMSPT3 and CPT.
For the CPT, we selected the span size through minimization of the AIC statistic. Many other methods of span selection are available. We believe similar results would be observed for any data driven span selection procedure, but further research is needed to confirm this. For the FMSPTs, we selected spans for a range across possible span sizes a priori. Other span sizes could be selected and power estimates may change accordingly. For the CPT and FMSPTs, we applied significance level adjustments based on empirical evidence from previous research and a nominal α level of 0.05 [27]. There is no guarantee that similar results will be observed in future studies as the significance cutoffs used here were selected and evaluated through a single set of simulations. For different nominal α levels, appropriate significance cutoffs must be determined. A number of extensions to the scan statistic are available, including elliptical [28] and flexibly shaped [29] zones; however for this research, our interest was in evaluating the original, and widely used, circular spatial scan statistic as applied using the software SaTScan. Applications of other versions of the scan statistic may influence the statistical power and sensitivity of the test. Evaluation of the extended methods is left for future research. In this research, we applied the methods to point data. Both the scan statistic and GAM methods are applicable to aggregate data and if applied to such data, the resulting distribution of power estimates would likely change.
Conclusions
Power of at least 80% indicates that the null hypothesis is correctly rejected at a high rate, a desirable quality of a testing method. The permutation tests each had power estimates exceeding the 80% threshold for large odds ratios. Reduced power was observed for a lower prevalence disease, as was expected with reduced theoretical power. The scan statistic had an observed power estimate of at least 80% for a circular cluster of increased risk centered in the study region but lower estimates for other variations in disease risk.
Sensitivities of at least 80% are desirable to ensure that the testing methods detect the correct areas of increased or decreased risk. In general, the FMSPTs had the highest sensitivity with estimates of at least 80% with large odds ratios for all disease risk patterns examined. The CPT had slightly lower sensitivity though its sensitivity reached 80% for higher prevalence diseases and with large odds ratios. The scan statistic had lower sensitivity estimates for all variations of disease risk examined and was observed to have at least 80% sensitivity only for a circular cluster centered in the study region which mimicked its own cluster detection method.
Simple patterns of spatial variation in disease risk were considered in this study. The relative pattern of power estimates of the four methods differ based on the pattern of disease risk considered. The spatial scan statistic outperformed the GAM methods in the case of a circular cluster centered in the study region, though it underperformed in sensitivity. For a linear association between geographic location and disease risk, the scan statistic had power estimates and sensitivity falling below the GAM estimates. It is important to note that analyses were performed using point data. Results of power comparisons applied to aggregate data may differ from those observed here. Across all simple scenarios examined in this research, the GAM methods presented a reasonable alternative with similar or greater power estimates and sensitivity exceeding that of the spatial scan statistic.
Declarations
Acknowledgements
This research was supported by grant P42ES007381 from the National Institute of Environmental Health (NIEHS), NIH. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NIEHS, NIH.
Authors’ Affiliations
References
 Marshall RJ: A Review of Methods for the Statistical Analysis of Spatial Patterns of Disease. Journal of the Royal Statistical Society, Series A (Statistics in Society) 1991, 154: 421–441.View Article
 Kulldorff M, Tango T, Park P: Power comparisons for disease clustering tests. Computational Statistics & Data Analysis 2003, 42: 665–684.View Article
 Takahashi K, Tango T: An extended power of cluster detection tests. Statistics in Medicine 2006, 25: 841–852.View ArticlePubMed
 Besag J, Newell J: The Detection of Clusters in Rare Diseases. Journal of the Royal Statistical Society, Series A (Statistics in Society) 1991, 154: 143–155.View Article
 Hastie TJ, Tibshirani RJ: Generalized Additive Models. New York: Chapman & Hall/CRC; 1990.
 Webster T, Vieira V, Weinberg J, Aschengrau A: Method for mapping populationbased casecontrol studies: an application using generalized additive models. International Journal of Health Geographics 2006., 5:
 Tusell F: A permutation test for randomness with power against smooth variation. Statistics and Computing 2001, 11: 147–154.View Article
 Kelsall JE, Diggle PJ: Spatial Variation in Risk of Disease: A Nonparametric Binary Regression Approach. Journal of the Royal Statistical Society Series C (Applied Statistics) 1998, 47: 559–573.View Article
 Hardle W, Huet S, Mammen E, Sperlich S: Bootstrap Inference In Semiparametric Generalized Additive Models. Econometric Theory 2004, 20: 265–300.View Article
 Cardinale M, Arrhenius F: The influence of stock structure and environmental conditions on the recruitment process of Baltic cod estimated using a generalized additive model. Canadian Journal of Fish and Aquatic Sciences 2000, 57: 2402–2409.View Article
 Young RL, Weinberg J, Vieira V, Ozonoff A, Webster TF: Generalized Additive Models and Inflated Type I Error Rates of Smoother Significance Tests. Computational Statistics & Data Analysis 2010, in press.
 Hurvich C, Simonoff J, Tsai CL: Smoothing Parameter Selection in Nonparametric Regression Using an Improved Akaike Information Criterion. Journal of the Royal Statistical Society, Series B (Methodological) 1998, 60: 271–293.View Article
 Kulldorff M, Nagarwalla N: Spatial Disease Clusters: Detection and Inference. Statistics in Medicine 1995, 14: 799–810.View ArticlePubMed
 Song C, Kulldorff M: Power evaluation of disease clustering tests. International Journal of Health Geographics 2003., 2:
 Ozonoff A, Bonetti M, Forsberg L, Pagano M: Power comparisons for an improved disease clustering test. Computational Statistics & Data Analysis 2005, 48: 679–684.View Article
 Ozonoff A, Webster T, Vieira V, Weinberg J, Ozonoff D, Aschengrau A: Cluster detection methods applied to the Upper Cape Cod cancer data. Environmental Health: A Global Access Science Journal 2005., 4:
 Kulldorff M: SaTScan. 7.0.3 edition. Boston and Maryland: Harvard Medical School, Boston and Information Management Services Inc, Silver Springs, Maryland; 2007.
 Aamodt G, Samuelsen SO, Skrondal A: A simulation study of three methods for detecting disease clusters. International Journal of Health Geographics 2006., 5:
 Aschengrau A, Weinberg J, Rogers S, Gallagher L, Winter M, Vieira V, Webster T, Ozonoff D: Prenatal Exposure to TetrachloroethyleneContaminated Drinking Water and the Risk of Adverse Birth Outcomes. Environmental Health Perspectives 2008, 116: 23–34.View Article
 Vieira V, Webster T, Weinberg J, Aschengrau A: Spatial analysis of bladder, kidney, and pancreatic cancer on upper Cape Cod: an application of generalized additive models to casecontrol data. Environmental Health Perspectives 2009., 8: Article 3
 Webster T, Vieira V, Weinberg J, Aschengrau A: Spatial analysis of casecontrol data using generalized additive models. In EUROHEIS/SAHSU Conference; 30–31 March 2003; Ostersund, Sweden Edited by: JL L. 2003, 56–59.
 Trepka MJ, Henrich J, Krause C, Schulz C, Lippold U, Meyer E, Wichmann HE: The Internal Burden of Lead among Children in a Smelter Town  A Small Area Analysis. Environmental Research 1997, 72: 118–130.View ArticlePubMed
 Wilhelm M, Ritz B: Residential Proximity to Traffic and Adverse Birth Outcomes in Los Angeles County, California, 1994–1996. Environmental Health Perspectives 2003, 111: 207–216.View ArticlePubMed
 McConnell R, Berhane K, Yao L, Jerrett M, Lurmann F, Gilliland F, Kunzli N, Gauderman J, Avol E, Thomas D, Peters J: Traffic, Susceptibility, and Childhood Asthma. Environmental Health Perspectives 2006, 114: 766–772.View ArticlePubMed
 Hastie TJ: gam: Generalized Additive Models. R Package 2008.
 R v 2.8.0 2.8.0 edition. The R Foundation for Statistical Computing; 2008.
 Young RL, Weinberg J, Vieira V, Ozonoff A, Webster T: The Power of Hypothesis Testing Using Generalized Additive Models with Bivariate Smoothers. 2009, in press.
 Kulldorff M, Huang L, Pickle L, Duczmal L: An elliptic spatial scan statistic. Statistics in Medicine 2006, 25: 3929–3943.View ArticlePubMed
 Tango T, Takahashi K: A flexibly shaped spatial scan statistic for detecting clusters. International Journal of Health Geographics 2005., 4:
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.