Thrombotic Thrombocytopenic Purpura Predominance in the Southern United States; Trends in Regional and Seasonal Variation and Mortality of TTP Utilizing the National Inpatient Sample Database Between 1999 and 2013

Abstract

Abstract Background: Few studies have been able to determine epidemiological factors associated with Thrombotic Thrombocytopenic Purpura (TTP) and no geographic distribution has been determined. Only one study to date, Giri et al., (Am J Hematol, 2015, 90:E24) has looked at regional differences in TTP and found a higher caseload of TTP in the South between 2009 and 2011. Data thus far on seasonal variations has been mixed. Park et al. (Transfusion, 2012, 52.7: 1530) reported an association between summer and presentation of TTP, and Giri et al. found a significant seasonal variation only in the Midwest with a spike in cases during the month of July, while others like Raval et al. (J Clin Apheresis, 2013, 29.2:113) reported no variation. Furthermore, mortality in TTP has traditionally been looked at in terms of data from a single center within their own regional cohort and no recent studies have analyzed the trends in mortality from TTP. Building on this, our study looked at the regional variation, seasonal variation, and national mortality of TTP using data on admissions from the National Inpatient Sample between 1999 and 2013. Methods: Adult admissions with a primary diagnosis of TTP were extracted from the National Inpatient Sample database using the ICD-9 code for TTP (446.2) during a 15 year period between 1999 and 2013 (N = 6,292, weighted N = 30,011). The sample was weighted to approximate the full inpatient population of the United States over the period of interest. Admission information studied included hospital region, which was differentiated into Northeast, Midwest or North Central, South, and West. The season of admissions over this interval was also analyzed. The Winter months were December to February, Spring: March-May, Summer: June-August, and Fall: September-November. We extrapolated seasons to be equivalent to climate. A Chi square analysis was used to analyze differences in categorical variables. To analyze differences in regional distribution, we assumed a normal distribution of 25% per region and then performed a chi squared analysis. Results: There were more cases of TTP in the South (43.50%) compared to the Midwest or North Central (23.31%), Northeast (18.36%), or West (14.82%) among 30,011 cases (p=0.001). In each year between 1999 and 2013, the admissions for TTP were higher in the South. While certain years individually appeared to have a seasonal variation, taking the years between 1999 and 2013 in summation (n=5,857) showed no statistically significant seasonal variation in presentation of TTP (p=0.694). Additionally, analyzing the seasonal data by region (n=5,857) still did not demonstrate any significant variation in the presentation of TTP (p=0.172). During the period of our analysis, overall mortality decreased from 12.12% in 1999 to its lowest level of 5.90% in 2013. Conclusion: These results not only confirm the findings from Giri et al., who showed a higher number of cases of TTP in the South for the years 2009 to 2011, but we also demonstrated this was true for the extended time period from 1999 to 2013. Our results also agree with Raval et al. who showed no seasonal association with presentation of TTP. Previous authors suggested that climate correlated with TTP admissions; however, our data showed that the climate did not have an effect on regional admissions for TTP. Extrapolating from this, even though the majority of cases of TTP occurred in the South, this is unrelated to the warmer climate seen in the South, as had been suggested by prior authors. From 1999 to 2013, hospital mortality trended downward. Perhaps trends in mortality have decreased due to recognition of effective treatment and earlier recognition of TTP. Further studies are needed to clarify why there is a regional difference in TTP and to determine if mortality will continue to decline over time. Disclosures No relevant conflicts of interest to declare.