Thyrotoxicosis: A Decade of Inpatient Trends in the United States

Presentation Number: MON 282
Date of Presentation: April 3rd, 2017

Skand Shekhar*1, Nilay Patel1, Apurva Bandekha2, Khushali Jhaveri3, Ekta Aneja4 and Anne Marie Van Hoven5
1Saint Peter's University Hospital, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, 2The Everette Clinic, Everett, WA, 3Mount Sinai Hospital, New York, NY, 4Saint Barnabas Hospital, Bronx, NY, 5Saint Peter's University Hospital, New Brunswick, NJ

Abstract

Background:

Thyrotoxicosis is defined as a clinical condition secondary to an excess amount of circulating thyroid hormone. The thyroid hormone may originate from either the thyroid gland or other sources. The population-based prevalence of clinical hyperthyroidism where thyroid hormone is derived from the thyroid gland, an important constituent of thyrotoxicosis, has been estimated to be around 0.5%. (1) However, no studies or data exist in the United States describing the population characteristics of patients admitted to US hospitals with this condition. Our study aimed to investigate the patient characteristics, geographical trends, comorbidities, length of stay (LOS), and hospitalization costs in these patients.

Methods:

We conducted a retrospective analysis of the National Inpatient Survey (NIS) database for a ten-year period between 2003-2012, using appropriate ICD-9 codes, with in hospital mortality as the end-point for patients aged 18 years or older. All analyses were performed using the designated weighting specified to the NIS data base to minimize bias. For categorical variables the χ2 test of trend for proportions was used with the Cochrane Armitage test via the “ptrend” command in Stata. Hierarchical two level logistic regression models were built to evaluate predictors of mortality.

Results:

We identified a total of 24871 hospitalizations that was extrapolated to a weighted number of 122159 hospitalizations with a primary diagnosis of thyrotoxicosis. The mean age of patients was 48.8 years. A majority of patients were females (77.6%). The overall inpatient mortality was 0.58% compared to a mortality of 10.1% according to a similar study in Japan. (2) Majority of these patients were treated at an urban setting (82%) and most of these patients were treated at large hospitals (67.17%). Geographically, southern US received at the largest share of these patients (38.6%) followed by the northeast (22%), the mid-west (20%) and the west (19%). The mean length of stay of these patients was 3.46 days, which declined from 3.7± 0.09 days in 2003 to 3.2±0.07 days in 2012. The mean cost of hospitalization was USD 7877, which increased from USD 7600±226 in 2003 to inflation adjusted USD 8191±184 in 2012. Factors significantly associated with higher mortality includes age in increments of 10 years (OR, CI and p-value: 1.19, 1.06-1.31, <0.01) and increased burden of comorbidities (CCI>=2; 3.34, 2.06-5.41, <0.001).

Conclusion:

Our study for the first time highlights the trends, disparities and similarities in the patient population admitted in hospitals. Moreover, increasing age and co-morbidity burden emerged as independent risk factors for poor prognosis. Further, over a period of 10 years, despite the fall in in hospital mortality and LOS, the adjusted health care cost increased which calls for investigation into the underlying reasons and corrective action.

 

Nothing to Disclose: SS, NP, AB, KJ, EA, AMV