nosed with IE . 7 Common causes of death in PWID include drug overdose , accidents , suicide , infection , and HIV-related mortality . 7 Thus , inadequate addiction care and ongoing IDU partially explain why long-term outcomes in PWID-IE are poor , and several studies show that addiction consultations in the treatment of PWID-IE may improve long-term survival in this population . 6 , 7
Poor long-term outcomes in PWID-IE can also be attributed to the high number of against medical advice ( AMA ) discharges seen in PWID-IE . 6 AMA discharges may result in serious infection due to receiving inadequate care . 6 Reasons for AMA discharges in these cases are multifold , including poor experiences with staff and inadequate pain management . 6 As such , it is imperative that providers address the root causes of AMA discharges to improve long-term outcomes in PWID-IE .
IE alone may specifically impact survival in PWID . A recent study comparing allcause mortality in PWID-IE and PWID hospitalized with other infections found that PWID with other infections had better 10-year survival than PWID-IE . 8 Straw et al .’ s findings suggest that cardiac infection may negatively impact survival in PWID beyond infection elsewhere . 8
There are significant healthcare costs associated with IE . An analysis of North Carolina hospital discharge data from 2010 to 2015 found that there was a twelvefold increase in hospitalizations for PWID-IE , and the average cost per hospitalization was over $ 50,000 . 4 The authors suggest that this indicates that the healthcare system and public payers could pay a larger proportion of the total cost of PWID-IE , which increased from $ 1.1 million in 2010 to $ 22.2 million in 2015 . 4 They assert that their findings suggest a need to focus on preventative strategies such as syringe exchange programs to reduce the morbidity and healthcare costs associated with PWID-IE . 4
In 2020 , West Virginia ( WV ) had the highest age-adjusted rate of drug overdose deaths involving opioids , reported as 68.6 deaths per 100,000 persons in 2020 . 9 This rate was nearly threefold higher than the national rate of 25.4 deaths per 100,000 persons in 2020 . 9 Local studies have produced findings consistent with these national trends , stating that cases of IE have increased concomitantly with mixed drug use throughout the last decade while placing an economic burden on the healthcare system . 10 However , it is worth noting that WV was one of six states to report a decrease in total fatal drug overdoses from 2021 to 2022 , which may be attributed to efforts across the state to destigmatize substance use disorder ( SUD ), expand recovery options , and increase naloxone
11 , 12 access .
Huntington , WV , is served by two hospitals that recently merged into one healthcare network . Although there are plans to create a data warehouse for the entire healthcare network , currently , only one hospital maintains a data warehouse for all patient records . Given the significant burden of IDU in our community , we sought to determine the number of cases of IE over time using the hospital data warehouse .
MATERIALS & METHODS
This study was conducted at Cabell Huntington Hospital ( CHH ), a 303-bed academic medical center located in Huntington , WV . We performed a retrospective review of electronic medical records for patients 12 years of age and older with an ICD-9 / ICD-10 diagnosis of IE ( n = 444 ) who were treated for initial or ongoing care within CHH between 2010 and 2019 . We started at age 12 years as this is at the younger end of the age range for reported PWID . 13 Potential cases were identified through both the Marshall Health Data Warehouse and CHH databases . This study received institutional review board approval from the Marshall University Office of Research Integrity ( IRB # 1487577 ).
All cases of IE were assigned a unique study identification number . Data variables regarding patient demographics , payer status , length of stay , hospital charges , readmissions , comorbid conditions , admitting vitals , laboratory results , hepatitis C virus ( HCV ) antibodies , and HIV screening results were obtained for each case . Data were abstracted electronically in a tab-separated variable format and were analyzed using Stata 17.0 ( College Station , TX ). All data provided to the investigators were de-identified using the Safe Harbor method and were stored on the CHH secured server with a password-protected computer system .
Echocardiographic data were collected on available patients ( n = 336 ). For subjects with available echocardiographic data , the following variables were collected : type of echocardiogram performed ( i . e ., TTE or TEE ), date of echocardiogram , presence of intracardiac mass or vegetation , and location of intracardiac mass or vegetation ( if present ).
Simple descriptive statistics were used to report on all variables in the study . Variables were analyzed using a bivariate analysis based on opioid use . A binary variable was created for opioid use ( no = 0 , yes = 1 ). Cases met the definition for opioid use if they contained ICD-9 / ICD-10 diagnoses of drug use history and / or positive opioid toxicology screenings . Cases that did not meet the criteria for opioid use were described as “ non-PWID-IE ” ( n = 172 ). Cases that met the criteria for opioid use were described as “ PWID-IE ” ( n = 272 ). Comparisons among variables were made using the Student ’ s t-test for continuous variables ( Wilcoxon Rank Sum if nonparametric ) or Chi-square for categorical variables . Continuous variables were described as medians and interquartile ranges ( IQRs ), and categorical variables were described as percentages . Tests of significance used a two-tailed alpha of 0.05 . The trend over time in IE cases by drug use was analyzed using Cuzick ’ s rank trend test . Odds ratios ( ORs ) with 95 % confidence intervals were used to further analyze the change in IE by PWID over time .
RESULTS
There were 444 cases with an IE diagnosis who were admitted to CHH between 2010 and 2019 . Based on our definition of opioid use , there were 172 cases of non- PWID-IE and 272 cases of PWID-IE . Annual admissions for IE are displayed in Figure 1 .
Cases meeting the criteria for opioid use ( PWID-IE ) ( n = 272 ) were observed more frequently than cases that did not meet
West Virginia Medical Journal • September 2024 • 15