Table 3. Clusters of US Census Block Groups With Substance Abuse Disorders Measured by Using the Discrete Poisson Model, New Orleans, 2004 and 2008.
| Cluster | Observed Hospitalizations | Expected Hospitalizations | RR | LLR | P Value |
|---|---|---|---|---|---|
| 2004 | |||||
| Most likely cluster, (New Orleans East) | 114 | 272.4 | 0.36 | 71.0 | <.001 |
| Secondary cluster 1, (Algiers, Bywater, Garden District, French Quarter, CBD) | 655 | 477.0 | 1.77 | 51.1 | <.001 |
| 2008 | |||||
| Most likely cluster, (Eastern corner of Mid-City) | 820.47 | 0.01 | 831.23 | 40.0 | <.001 |
| Secondary cluster 1, (Lower 9th Ward) | 11.60 | 1.29 | 11.91 | 23.2 | <.001 |
| Secondary cluster 2, (Algiers) | 0.44 | 96.86 | 0.40 | 22.1 | <.001 |
| Secondary cluster 3, (Southeast corner of Mid-City) | 23.14 | 0.26 | 23.38 | 13.1 | <.001 |
Abbreviations: RR, relative risk; LLR, log-likelihood ratio; CBD, Central Business District