Journal of Pediatric Nursing (2015) 30, 346–352

Hospitalization Cost per Case of Neonatal Herpes Simplex Virus Infection From Claims Data1,2,3,4 Kwame Owusu-Edusei Jr. PhD ⁎, Elaine W. Flagg PhD, Thomas L. Gift PhD Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA Received 17 April 2014; revised 7 August 2014; accepted 8 August 2014

Key words: Herpes simplex virus; Neonates; Inpatient cost; Insured population

The purpose of this study was to estimate the average excess inpatient cost of neonatal herpes simplex virus (NHSV) infection from 2005 to 2009 insurance claims data. The estimated adjusted average excess inpatient cost for neonate admissions with HSV diagnosis and N 7 days of hospitalization was $40,044 [95% confidence interval (CI), $33,529–$47,775]. When disaggregated by the days of admission, cost estimates were: 8–13 days, $23,918 [CI, $19,490–$29,282]; 14–21 days, $44,358 [CI, $34,654–$56,673]; N 21 days, $68,916 [CI, $49,905–$94,967]). Although these estimates are not representative of the entire US, they can inform future economic evaluation studies on NHSV interventions. Published by Elsevier Inc.

NEONATAL HERPES SIMPLEX virus (NHSV) infection is typically caused by the perinatal transmission of herpes simplex virus, most frequently through contact with herpes simplex virus (HSV) infected genital secretions from the mother at the time of delivery (Bradford, Whitley, & Stagno, 2008; Lawrence Corey & Wald, 2008). Mothers who acquire HSV during pregnancy have a substantially higher risk of transmission than mothers with established disease (Lawrence Corey & Wald, 2008; L. Corey & Wald, 2009). However, a large proportion (over 70%) of HSVinfected neonates are born to mothers without symptoms or signs of HSV lesions at the time of delivery (Corey & Wald, 2008), making detection very challenging. In the US, reporting of NHSV cases is not required and we are not aware of any trend analysis studies of the burden of NHSV. Therefore, it is not known whether the incidence of

1

Funding: None. Conflicts of interest: None. 3 Previous presentation: A version of the results from this study was presented as a poster at the National STD Prevention Conference in Minneapolis, Minnesota, 12–15 March, 2012. 4 Commercial financial support: None. ⁎ Corresponding author: Kwame Owusu-Edusei Jr. PhD. E-mail address: [email protected]. 2

http://dx.doi.org/10.1016/j.pedn.2014.08.004 0882-5963/Published by Elsevier Inc.

disease has changed over time. The estimated rate of NHSV among live births in 2006 was 9.6 per 100,000 (Flagg & Weinstock, 2011). Most incidence estimates from managed care populations and from population-based state and local studies have ranged from 8.4 to 28.2 per 100,000 births (Flagg & Weinstock, 2011). Incidence data for NHSV are similar to those for perinatal human immunodeficiency virus (HIV) before the advent of routine use of antiretroviral therapy during pregnancy, and the incidence is similar to, or higher than, that of congenital syphilis, toxoplasmosis, and congenital rubella in years in which the virus was not epidemic (Corey & Wald, 2009). A case series of 8538 pregnant women receiving care in two hospitals from 1989 to 1993 used serum samples obtained at the first prenatal visit and at the time of labor and found that 24% were HSV-negative at entry. Of the 7046 women who were HIV-negative at entry, 1.3% became seropositive for either HSV-1 or HSV-2 (Brown et al., 1997). In a recent study, 24.5% of cases identified in 2006 were admitted for inpatient stay with NHSV diagnosis at birth, and 86.9% of cases identified were admitted for inpatient stay after birth, but within 30 days of birth (Flagg & Weinstock, 2011). In newborns, HSV infection can manifest as (1) disseminated disease involving multiple organs in about 25% of cases, (2) localize central nervous system (CNS) disease in about

Cost per Case of Neonatal Herpes Simplex Virus Infection 30% of cases, or (3) disease localized to the skin, eyes and/or mouth (SEM) in about 45% of cases (American Academy of Pediatrics, 2012). Although the majority (approximately twothirds) of neonates with disseminated or CNS disease have skin lesions, the lesions may not be present at the time of initial presentation. Thus, the diagnosis of NHSV is difficult in the absence of lesions. Consequently, disseminated infection should be considered in neonates with sepsis syndrome, negative bacteriologic culture results, and severe liver dysfunctions (American Academy of Pediatrics, 2012). NHSV should also be considered in neonates with fever, irritability, and abnormal cerebrospinal fluid (CSF) findings, especially in the presence of seizures or during a time of year when enteroviruses are not circulating in the community. Although asymptomatic HSV infection is common in older children, it rarely, if ever, occurs in neonates (American Academy of Pediatrics, 2012). Studies have reported that the mortality rates among untreated severe cases can range from 50 to 85% (Whitley, Nahmias, Visintine, Fleming, & Alford, 1980; Whitley et al., 1980). The test for NHSV comprises the isolation of HSV by viral culture from specimen obtained from skin vesicles and other sites (such as CSF, stool, urine, throat, nasopharynx and conjuctivae). Additionally, the detection of HSV-DNA in CSF by polymerase chain reaction (PCR) is the typical diagnostic method for CNS (American Academy of Pediatrics, 2012; Bradford et al., 2008). However, in administrative claims data, International Classification of Diseases, Clinical Modification (ICD-CM) codes for herpes simplex virus infection in neonates indicate NHSV diagnosis (Flagg & Weinstock, 2011). The recommended regimen for NHSV involves 14 to 21 days of intravenous acyclovir (Workowski & Berman, 2010). Thus, besides the pain and suffering endured by neonates infected with HSV, the treatment can be very costly due in part to the relatively longer admission time required for treatment. However, estimates of the hospitalization cost due to NHSV in the US are limited. In this study, we estimate the excess/additional inpatient cost attributable to HSV diagnosis in neonates using US insurance claims data for 2005–2009.

Materials and Methods Data Source and Case Definition Claims data on inpatient admission from the Truven Health Analytics MarketScan® Commercial Claims and Encounters Database (Truven Health Analytics, Ann Arbor, MI) for 2005 through 2009 were assessed and used in this study. In 2005, the MarketScan database contained claims data on over 17 million people who had employer-sponsored health insurance from more than one hundred payers, including large employers, health plans, government and public organizations (Truven Health Analytics, 2011). The records comprise information

347 on fully adjudicated and paid claims for persons with employer-sponsored health plans, including their spouses and dependents (Adamson, Chang, & Hansen, 2006). Details on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for single or multiple liveborn infants who were identified as “healthy/uncomplicated normal newborns” is provided elsewhere (Owusu-Edusei, Introcaso, & Chesson, 2013). We used ICD-9-CM codes 054.0–054.9 (Buck, 2011) to identify admissions with a HSV diagnosis. Liveborn admissions that were not coded as “normal newborns” and which did not have a HSV diagnosis were excluded from our analyses. Also, because codes for fever or seizures can be symptoms of other viral or bacterial infections (Caviness, Demmler, Swint, & Cantor, 2008), neonates with codes for fever or seizure, but without ICD-9-CM codes for herpes simplex infection (054.0–054.9) were excluded.

Estimating Total and Excess Inpatient Costs Attributable to NHSV In the MarketScan database, inpatient total costs were the total gross payments to all providers who submitted claims for covered services rendered during admission (Truven Health Analytics, 2011). Records with total payments less than $1 were deleted because most were likely due to data entry errors, although we found very few cases where the total payment was less than $1 (b 0.1%). We first estimated total costs associated with NHSV without regard to length of hospital admission (i.e., length of stay). Then, following Flagg and Weinstock (2011), we estimated total NHSV cost for those with length of stay N 7 days. Finally, we re-estimated the hospitalization costs for NHSV by disaggregating the estimated cost based on the length of stay as follows: 8–13, 14–21 and N 21 days (Caviness et al., 2008). Following Owusu-Edusei et al. (2013), we computed simple summary statistics on the total payments for those with and without HSV diagnosis to determine excess costs (referred to as unadjusted estimates). Second, following previous studies (Bateman, Phibbs, Joyce, & Heagarty, 1997; Owusu-Edusei et al., 2013; Shrestha, Zhang, Albright, & Imperatore, 2011), we transformed the total costs into natural logarithm (to reduce the influence of outliers) of total costs and applied a semi-log regression model (i.e., the dependent variable was the natural logarithm of total cost) to determine costs while controlling for enrollee characteristics (i.e., independent variables) provided in the database (referred to as adjusted estimates). This approach was used because the total inpatient admission costs are largely dependent on (or are determined by) the patient and insurance characteristics provided in the claim record. Thus, we investigated the available characteristics that were influential (statistically significant) in determining the total inpatient cost. Then, we used the regression results to determine the adjusted cost—the mean inpatient cost after controlling for the available characteristics.

348

K. Owusu-Edusei Jr. et al.

Table 1 Summary characteristics of neonates with and without herpes simplex virus (HSV) diagnosis from privately-insured claims data, 2005–2009 a. Variable

Unadjusted cost (2009 US dollars) Length of stay (days) Gender Male Female Data source Employer Health plan Drug benefits No Yes Year of service 2005 2006 2007 2008 2009 Region Northeast North Central South West Unknown Health plan Comprehensive Exclusive provider organization (EPO) Health maintenance organization (HMO) Point of service (POS) Preferred provider organization (PPO) Point of service (POS) with capitation Consumer driven health plan (CDHP) High deductible health plan (HDHP)

HSV diagnosis (n = 150)

HSV diagnosis and admission N 7 days (n = 73)

Healthy neonates (n = 474,743)

Mean (standard deviation)

Mean (standard deviation)

Mean (standard deviation)

32,197 (47,571) 10.8 (11.5) Frequency (%) 83 (55) 67 (45)

55,433 (59,337) 18.5 (12.5) Frequency (%) 44 (60) 29 (40)

1681 (4168) 2 (1.3) Frequency (%) 237,130 (50) 237,613 (50)

67 (45) 83 (55)

36 (49) 37 (51)

259,074 (55) 215,669 (45)

34 (23) 116 (77)

15 (21) 58 (79)

75,955 (16) 398,743 (84)

17 14 38 29 52

(11) (9) (25) (19) (35)

6 (8) 8 (11) 15 (21) 19 (26) 25 (34)

58,406 (12) 62,024 (13) 109,025 (23) 111,737 (24) 133,551 (28)

17 (11) 35 (23) 74 (49) 22 (15) 2 (1)

6 (8) 19 (26) 39 (53) 8 (11) 1 (1)

56,552 (12) 151,563 (32) 190,390 (40) 68,362 (14) 7876 (2)

5 (3) 2 (1) 21 (14) 12 (8) 106 (71) 2 (1) 2 (1) 0

2 (3) 1 (1) 11 (15) 9 (12) 47 (64) 2 (3) 1 (1) 0

26,522 (6) 3924 (1) 73,540 (15) 52,412 (11) 303,571 (64) 2525 (1) 9497 (2) 1935 (0)

HSV, herpes simplex virus. a Percentages may not add up to 100 due to rounding.

The control/independent variables were dichotomous or polychotomous variables that included gender, source of data (i.e., self-insured employer or health plan), whether a drug benefit was identified for the patient, year of service, region, health plan type [such as health maintenance organization (HMO) and provider preferred organization (PPO)], employment classification (such as salary and non-salary) and employment status (such as active full-time and active seasonal). Additionally, we created variables for neonate patients with HSV diagnosis: (1) a dichotomous variable for HSV diagnosis equal to 1 if diagnosed with HSV, zero otherwise and (2) a dichotomous variable (equal to 1, zero otherwise) for those with HSV diagnosis whose length of stay was N 7 days, and (3) dichotomous variables (equal to 1, zero otherwise) for HSV diagnosis with lengths of stays ranging from 8 to 13, 14 to 21 and over 21 days.

We explored a state-specific fixed-effects model to account for variations in costs within and across states in our preliminary analyses. However, when we included the regional dichotomous variables, the state-specific fixedeffect variables were dropped due to high multicollinearity. To obtain robust standard errors and confidence intervals, we computed bootstrapped standard errors for the coefficients from 50 estimations (Kennedy, 2008). Given the natural log transformation of the dependent variable (i.e., the total costs), the regression coefficients of the independent variables were converted as (expcoefficient − 1), and interpreted as the relative difference in the adjusted mean costs when compared to the adjusted mean cost for the referent group—healthy neonates (Kennedy, 2008). As an example, a coefficient of 2 implies that the adjusted excess cost for the specified group was 6.4 (i.e., exp2 − 1 = 6.3891) times that of the adjusted mean total cost for the healthy neonates.

Cost per Case of Neonatal Herpes Simplex Virus Infection Details of all the software applications used for this study are presented elsewhere (Owusu-Edusei et al., 2013). To make costs comparable across years, we adjusted all cost data to 2009 US dollars using the medical care component of the Consumer Price Index for All Urban Consumers (United States Department of Labor, 2011).

Results A summary of our results from the extracted data, including relevant enrollee characteristics and unadjusted cost is presented in Table 1. Based on the available inpatient admissions data from 2005 to 2009, we identified 150 neonate admission claims with HSV diagnosis and almost half a million uncomplicated normal neonate claim records (n = 474,743) that were usable (i.e., after deleting those with missing entries or miscoded variables). Approximately 49% (n = 73) of neonates with a HSV diagnosis were hospitalized for more than 7 days (Table 1). The estimated unadjusted mean inpatient admission cost was $1681 for uncomplicated healthy neonates, $32,197 for neonate admission with HSV diagnosis and $55,433 for those with HSV whose length of stay was N 7 days. The disaggregated estimated unadjusted costs were as follows: 8– 13 days, $28,139; 14–21 days, $53,458; N 21 days, $95,692. Thus, the excess unadjusted hospitalization costs were $30,513 for admissions with HSV diagnosis and $53,752 for those with HSV diagnosis and N 7 days stay in the hospital (8–13 days, $26,458; 14–21 days, $51,777; N 21 days, $94,011). The mean length of stay was 2 days for uncomplicated healthy neonate admissions, almost 11 days for admissions with HSV diagnoses and 18.5 days for those with length of stay N 7 days (Table 1). Our results from the extracted data indicated that the majority of patients with HSV diagnosis were males and over three quarters (N 75%) had drug prescription benefits (Table 1). The number of patients with HSV diagnosis was 17 (11%) in 2005 and 14 (9%) in 2006. However, from 2007 to 2009, the number of patients with HSV diagnosis increased, rising to 38 (25%) and 52 (35%), respectively. Almost half (49%) of the identified HSV diagnoses were from the Southern region of the US. Additionally, the proportion of the neonates that had HSV diagnosis claims from the Southern region was about 9% higher than the overall proportion of healthy neonates (49 vs. 40%) from the Southern region (Table 1). In contrast, the proportion of neonates in our dataset that had HSV diagnosis claims from each region was lower or about the same as the proportion of healthy neonates from each region (Table 1). The dominant health plan for neonate enrollees was the preferred provider organization (PPO) which represented over 64% of the health plans for all neonates (Table 1). Over 88% of neonates diagnosed with HSV were discharged to go home; however, 2 patients (1%) were transferred to shortterm hospitals for further medical care and 2 (1%) died.

349

Regression Analyses A summary of our regression results is presented in Table 2. Model 1 included the dichotomous variable for patients with HSV diagnosis, while model 2 included the dichotomous variable for HSV admissions N 7 days. Model 3 included the disaggregated dichotomous variables for those with HSV diagnosis whose length of stay was N 7 days (i.e., 8–13, 14–21 and N 21 days). All the models included two additional polychotomous variables (i.e., 9 employment classification categories and 9 employment status categories); however, we did not include the results for these variables in Table 2 (detailed results of the regression are available from the lead author). Based on the regression results, the adjusted mean cost of inpatient admission was $1408 and the estimated adjusted excess inpatient cost for those with HSV diagnosis (regardless of length of stay) was 10.21 times (i.e., exp2.4166 − 1; p b 0.01) that of the inpatient cost of healthy neonate admissions. However, among those with HSV diagnosis, those who stayed for N 7 days incurred 28.44 times [i.e., exp3.3823 − 1; (p b 0.01)] that of the mean inpatient cost for healthy neonate admissions. Applying similar transformations to the estimated disaggregated coefficients, the excess costs for those with 8–13, 14–21 and N 21 days of stay were 16.99 (p b 0.01), 31.50 (p b 0.01) and 48.94 (p b 0.01) times the cost of the healthy neonate admission, respectively (Table 2). These estimates implied that the excess hospitalization costs associated with NHSV were: $14,373 [95% confidence interval (CI), $11,674–$17,747], and $40,044 [CI, $33,529–$47,775] for those with HSV diagnosis and N 7 days of hospitalization. The disaggregated costs were: 8–13 days, $23,918 [CI, $19,490–$29,282]; 14– 21 days, $44,358 [CI, $34,654–$56,673]; N 21 days, $68,918 [CI, $49,905–$94,967].

Summary of Results A complete summary of all the cost estimates (adjusted and unadjusted) is provided in Table 3 for easy comparison. As expected, the adjusted estimates were substantially lower than the unadjusted estimates, due largely to the reduced influence of outliers on the estimates for each category. Also as expected, the length of stay had a substantial impact on the hospitalization cost—costs increased with length of stay (Table 3).

Control Variables Our results from the regression analyses indicated that the mean inpatient admission cost for healthy female neonates was approximately 13% (p b 0.01) lower than the mean admission cost for healthy male neonates (Table 2). The mean inpatient cost for healthy neonates was significantly lower (8%, p b 0.01) if the data were provided by a health plan than when the data were provided by a self-insured

350

K. Owusu-Edusei Jr. et al.

Table 2 Summary Results From Regression Analyses Used to Estimate the Excess Inpatient Admission Cost Associated With Herpes Simplex Virus Diagnosis in Neonates Using Privately-Insured Claims Data, 2005–2009: Dependent Variable, Natural Log of Total Inpatient Cost (n = 474,893). ⁎ Variable Cost and length of stay Healthy neonate HSV Dx (n = 150) HSV Dx, N 7 days (n = 73) HSV Dx, 8–13 days (n = 26) HSV Dx, 14–21 days (n = 28) HSV Dx, N 21 days (n = 19) Gender Male Female Data source Employer Health Plan Drug benefits identified? No Yes Year of service 2005 2006 2007 2008 2009 Region Northeast North Central South West Unknown Health plan Comprehensive EPO HMO POS PPO POS with capitation CDHP HDHP

Model 1 (95% CI)

Model 2 (95% CI)

Model 3 (95% CI)

Referent 2.4166 (2.2229 to 2.6103) †

Referent

Referent

3.3823 (3.2113 to 3.5533) †

2.8896 (2.6974 to 3.0817) † 3.4813 (3.2430 to 3.7196) † 3.9109 (3.5957 to 4.2260) †

Referent Referent Referent − 0.1365 (− 0.1398 to − 0.1332) † − 0.1364 (− 0.1393 to − 0.1335) † − 0.1364 (− 0.1394 to − 0.1333) † Referent Referent Referent − 0.0802 (− 0.0879 to − 0.0726) † − 0.0802 (− 0.0884 to − 0.0719) † − 0.0802 (− 0.0879 to − 0.0724) † Referent 0.0532 (0.0474 to 0.0589) †

Referent 0.0531 (0.0484 to 0.0578) †

Referent 0.0531 (0.0491 to 0.0571) †

Referent 0.0953 (0.0885 0.0463 (0.0404 0.0620 (0.0560 0.1529 (0.1470

Referent 0.0950 (0.0874 0.0463 (0.0389 0.0617 (0.0556 0.1528 (0.1459

Referent 0.0950 (0.0888 to 0.1012) † 0.0463 (0.0407 to 0.0519) † 0.0617 (0.0558 to 0.0676) † 0.1528 (0.1466 to 0.1590) †

to 0.1020) † to 0.0523) † to 0.0680) † to 0.1587) †

to 0.1027) † to 0.0539) † to 0.0677) † to 0.1596) †

Referent − 0.2102 (− 0.2151 to − 0.2053) † − 0.4100 (− 0.4151 to − 0.4050) † − 0.3388 (− 0.3466 to − 0.3310) † − 0.2872 (− 0.3052 to − 0.2691) †

Referent − 0.2105 (− 0.2151 to − 0.2060) † − 0.4102 (− 0.4151 to − 0.4054) † − 0.3389 (− 0.3460 to − 0.3318) † − 0.2872 (− 0.3007 to − 0.2737) †

Referent − 0.2105 (− 0.2166 to − 0.2045) † − 0.4102 (− 0.4163 to − 0.4042) † − 0.3389 (− 0.3459 to − 0.3318) † − 0.2873 (− 0.3045 to − 0.2700) †

Referent − 0.0414 (− 0.0590 to − 0.0238) † − 0.0617 (− 0.0716 to − 0.0517) † − 0.0620 (− 0.0716 to − 0.0524) † − 0.0586 (− 0.0673 to − 0.0450) † 0.0044 (− 0.0177 to 0.0266) − 0.0001 (− 0.0124 to 0.0122) − 0.0157 (− 0.0397 to 0.0082)

Referent − 0.0412 (− 0.0594 to − 0.0230) † − 0.0616 (− 0.0710 to − 0.0523) † − 0.0622 (− 0.0723 to − 0.0522) † − 0.0585 (− 0.0674 to − 0.0496) † 0.0034 (− 0.0217 to 0.0285) 0.0001 (− 0.0136 to 0.0135) − 0.0157 (− 0.0415 to 0.0097)

Referent − 0.0411 (− 0.0613 to − 0.0209) † − 0.0617 (− 0.0717 to − 0.0517) † − 0.0623 (− 0.0712 to − 0.0534) † − 0.0585 (− 0.0670 to − 0.0501) † 0.0031 (− 0.0200 to 0.0262) 0.0001 (− 0.0135 to 0.0132) − 0.0159 (− 0.0396 to 0.0077)

Note. CI = 95% confidence interval, HSV = neonatal herpes simplex virus, Dx = diagnosis, EPO = exclusive provider organization, HMO = health maintenance organization, POS = point of service, PPO = preferred provider organization, CDHP = consumer-driven health plan, HDHP = high deductible health plan. ⁎ All models included employment classification and employment status polychotomous variables. † Indicates P b .01, calculated using bootstrapped generated standard errors.

employer. Also, the mean inpatient cost for healthy neonates was significantly higher (5%, p b 0.01) for those with drug benefits than for those without drug benefits (Table 2). Compared to inpatient costs in 2005, the mean costs for healthy neonates in calendar years 2006–2009 were at least 5% (p b 0.01) higher. The mean inpatient costs for healthy neonates from the Northeast region was significantly higher (at least 19%, p b 0.01) than the mean inpatient cost from the other regions (Table 2). When we made the South the referent region, our results indicated that the mean inpatient cost for healthy neonates in the other regions were at least 7% (p b 0.01) higher than those from the South (not shown).

Discussion This study used inpatient admission claims data from 2005 to 2009 to identify and estimate the excess hospitalization cost attributable to NHSV diagnosis among a large cohort of US neonates with private health insurance. We identified 150 NHSV inpatients and almost half (49%; n = 73) of them were hospitalized for N 7 days. Based on our case definition, which included HSV diagnosis and N 7 days of hospitalization, we estimated that the mean excess hospitalization cost of NHSV (when compared with the cost for healthy neonates) was $40,044 [CI, $33,529–

Cost per Case of Neonatal Herpes Simplex Virus Infection Table 3

351

Summary hospitalization cost (2009 US dollars) for neonates with and without HSV diagnosis.

Health status Healthy neonate Neonate with HSV Neonate with HSV Neonate with HSV Neonate with HSV Neonate with HSV

Dx (excess cost) Dx, N 7 days (excess cost) Dx, 8–13 days (excess cost) Dx, 14–21 days (excess cost) Dx, N 21 days (excess cost)

Unadjusted cost [95% (CI)]

Adjusted cost [95% (CI)]

1681 [1670–1693] 30,513 [22,852–38,179] 53,752 [39,918–67,584] 26,458 [20,800–32,103] 51,777 [38,757–64,796] 94,011 [47,480–140,541]

1408 [1399–1416] 14,373 [11,674–17,747] 40,044 [33,529–47,775] 23,918 [19,490–29,282] 44,358 [34,654–56,673] 68,918 [49,905–94,967]

HSV, herpes simplex virus; CI, confidence interval; Dx, diagnosis.

$47,775]. When we estimated cost for all neonates with HSV diagnosis, the excess cost was $14,373 [CI, $11,674– $17,747]. However, it is possible that neonates with HSV admitted for less than 8 days were those for whom an HSV diagnosis was ruled out, or who were discharged and subsequently treated for HSV on an outpatient basis (Caviness et al., 2008; Flagg & Weinstock, 2011). Our disaggregated costs which were based on ranges of the length-of-stay were higher than those reported in a recent cost-effectiveness study (Caviness et al., 2008). Additionally, Flagg and Weinstock (2011) reported a substantially higher mean hospitalization cost of $92,664 in 2006 US dollars. However, this overestimated the hospitalization cost of HSV because it was not adjusted by the existing cost-to-charge ratios (cost-to-charge ratios are used to estimate the resource cost of inpatient care and its variation across hospitals and conditions Healthcare Cost and Utilization Project, 2012). Our estimate of the cost of hospitalization for uncomplicated normal births (healthy neonates) was consistent with findings from a recent cost analysis study of neonate care in California (Schmitt, Sneed, & Phibbs, 2006). Additionally, our estimated mean number of days spent in the hospital for uncomplicated neonate admissions (2 days) was consistent with the 1.9 days reported by a recent study (Russell et al., 2007). Our finding that the cost of inpatient admissions was lowest for the South is consistent with a recent study of the cost of congenital syphilis from claims data in the US (Owusu-Edusei et al., 2013). Consistent with our expectation, we found a lower (13%) admission cost for females than for males. Other studies have shown a high rate of male circumcision—over 55% (Centers for Disease, C., & Prevention, 2011)—performed within the first 2 days of birth in the US, which represents an additional cost not incurred in female live births. We found that the largest proportion of the neonates in our dataset was from the Southern region. The disproportionate number of cases we found in the Southern region is similar to findings from a previous study of congenital syphilis using claims data (Owusu-Edusei et al., 2013), and consistent with the burden of other sexually transmitted infections in the US (Centers for Disease Control and Prevention, 2013). We found that the number of inpatient claims with HSV diagnosis increased from 2005 to 2009. This may be attributable to the increase in the number of enrollees

represented in the database over the years. The number of enrollees more than doubled from 17 million in 2005 to almost 40 million in 2009 (Truven Health Analytics, 2011). Our study has several limitations that are worth noting. First, administrative health claims data are not perfect. Diagnosis codes in claims can be inaccurate owing to data entry errors or miscoding (Henderson, Shepheard, & Sundararajan, 2006; Owusu-Edusei et al., 2013; Peabody, Luck, Jain, Bertenthal, & Glassman, 2004). In addition, as discussed above, diagnosis codes may be ruled-out but retained in the claims data (Flagg & Weinstock, 2011). Our case definition (HSV diagnosis during birth admission with admission N 7 days) differs from that used by Flagg and Weinstock (2011) (HSV diagnosis within the first 60 days of life with admission N 7 days) due to limitations in the way age is coded in the MarketScan data. Thus, our estimate is for neonates b 1 year old. Flagg and Weinstock found that only 25% of neonatal HSV cases were admitted at birth. It is possible that cases of NHSV admitted at birth differ from those diagnosed after the birth admission, but within 60 days, and in terms of severity of disease and time until definitive diagnosis and treatment initiation, but the nature and extent of these differences are not known. Our cost estimates cannot be generalized because they are derived using a convenience sample of neonates who were insured through large employer-sponsored health care programs in the US. Also, our estimates do not include other costs incurred before admission and after discharge, the cost of long-term sequelae from HSV infection, or associated outpatient visit costs. Additionally, our estimates do not include other costs not covered by insurance, intangible costs such as pain and suffering, or productivity losses incurred by family members. Finally, another major limitation worth noting is the fairly small sample size of the NHSV cases we identified in this study. Thus, even though the excess inpatient costs by admission days might be significantly different across regions, the small sample size precluded us from further analyses to assess regional differences (or similarities). Our study has several strengths, as well. The MarketScan database is one of the largest private insurance databases in the US. Secondly, and most important, the costs provided in the database are payments (not hospital charges), so our estimate represents actual dollar amounts paid for the medical services delivered (Adamson et al., 2006; Truven Health Analytics, 2011). Finally, to our knowledge, this is the only

352 study that focused entirely on the estimation of the total direct inpatient cost of NHSV.

Conclusion For researchers and health care analysts, the estimates provided in this study present additional information that can help to quantify the overall economic burden of NHSV infections, as well as inform the economic evaluation of future interventions to prevent and/or reduce NHSV infections in the US. Additionally, when there are equally effective treatment/ preventive alternatives for NHSV control and prevention, cost considerations become very critical in the decision-making process for health policy makers, health care providers as well as patients. Given the lack of information on NHSV, more research is needed to understand the differences in NHSV presentation, treatment and associated outcomes, especially for those diagnosed at birth or within the first 60 days of life.

References Adamson, D. M., Chang, S., & Hansen, L. G. (2006 Health research data for the real world: The MarketScan databases. White Paper. Thomson Healthcare, 1–32. American Academy of Pediatrics (2012 Herpes simplex: Clinical manifestations. In L. K. Pickering (Ed.), Red Book: 2012 report of the committee on infectious diseases (29th ed.). Elk Grove Village, IL: American Academy of Pediatrics. Bateman, D. A., Phibbs, C. S., Joyce, T., & Heagarty, M. C. (1997 The hospital cost of congenital syphilis. Journal of Pediatrics, 130, 752–758. Bradford, R. D., Whitley, R. J., & Stagno, S. (2008 Herpes virus infection in neonates and children: Cytomegalovirus and herpes simplex virus. In K. K. Holmes, P. F. Sparling, W. E. Stamm, P. Piot, J. N. Wasserheit, L. Corey, M. S. Cohen, & H. D. Watts (Eds.), Sexually transmitted diseases (pp. 1629–1658). New York, New York: McGraw Hill. Brown, Z. A., Selke, S., Zeh, J., Kopelman, J., Maslow, A., Ashley, R. L., et al. (1997 The acquisition of herpes simplex virus during pregnancy. New England Journal of Medicine, 337, 509–515, http://dx.doi.org/10. 1056/NEJM199708213370801. Buck, C. J. (2011 2011 ICD-9-CM for hospitals, volumes 1, 2 & 3 (professional ed. vol. 1, 2 & 3). St Louis, Missouri: Elsevier Saunders. Caviness, A. C., Demmler, G. J., Swint, J. M., & Cantor, S. B. (2008 Costeffectiveness analysis of herpes simplex virus testing and treatment strategies in febrile neonates. Archives of Pediatrics and Adolescent Medicine, 162, 665–674, http://dx.doi.org/10.1001/archpedi.162.7.665.

K. Owusu-Edusei Jr. et al. Centers for Disease Control and Prevention (2013 Sexually transmitted disease surveillance, 2012. Atlanta, GA: U.S. Department of Health and Human Services. Centers for Disease, C., & Prevention (2011 Trends in in-hospital newborn male circumcision—United States, 1999–2010. MMWR. Morbidity and Mortality Weekly Report, 60, 1167–1168. Corey, L., & Wald, A. (2008 Genital Herpes. In K. K. Holmes, P. F. Sparling, W. E. Stamm, P. Piot, J. N. Wasserheit, L. Corey, M. S. Cohen, & H. D. Watts (Eds.), Sexually transmitted diseases (pp. 399–429). New York: McGraw Hill. Corey, L., & Wald, A. (2009 Maternal and neonatal herpes simplex virus infections. New England Journal of Medicine, 361, 1376–1385, http://dx.doi.org/10.1056/NEJMra0807633. Flagg, E. W., & Weinstock, H. (2011 Incidence of neonatal herpes simplex virus infections in the United States, 2006. Pediatrics, 127, e1–e8, http://dx.doi.org/10.1542/peds.2010-0134. Healthcare Cost and Utilization Project (2012 Cost-to-charge ratio files. Retrieved December, 2012, from. http://www.hcup-us.ahrq.gov/db/ state/costtocharge.jsp Henderson, T., Shepheard, J., & Sundararajan, V. (2006 Quality of diagnosis and procedure coding in ICD-10 administrative data. Medical Care, 44, 1011–1019. Kennedy, P. (2008 A guide to econometrics (6th ed.). Malden, Massachussets: Blackwell Publishing. Owusu-Edusei, K., Jr., Introcaso, C. E., & Chesson, H. W. (2013 Hospitalization cost of congenital syphilis diagnosis from insurance claims data in the United States. Sexually Transmitted Diseases, 40, 226–229. Peabody, J. W., Luck, J., Jain, S., Bertenthal, D., & Glassman, P. (2004 Assessing the accuracy of administrative data in health information systems. Medical Care, 42, 1066–1072. Russell, R. B., Green, N. S., Steiner, C. A., Meikle, S., Howse, J. L., Poschman, K., et al. (2007 Cost of hospitalization for preterm and low birth weight infants in the United States. Pediatrics, 120, e1–e9 [doi: 120/1/e1 [pii] 10.1542/peds.2006-2386]. Schmitt, S. K., Sneed, L., & Phibbs, C. S. (2006 Costs of newborn care in California: A population-based study. Pediatrics, 117, 154–160, http://dx.doi.org/10.1542/peds.2005-0484. Shrestha, S. S., Zhang, P., Albright, A., & Imperatore, G. (2011 Medical expenditures associated with diabetes among privately insured U.S. youth in 2007. Diabetes Care, 34, 1097–1101, http://dx.doi.org/10.2337/dc10-2177. Truven Health Analytics (2011 Marketscan research database: User guide and database dictionary. Ann Arbor, MI: Truven Health Analytics. United States Department of Labor (2011 Consumer price indexes—All urban consumers. Retrieved December 15, 2013, from. http://www.bls. gov/cpi/home.htm Whitley, R. J., Nahmias, A. J., Soong, S. J., Galasso, G. G., Fleming, C. L., & Alford, C. A. (1980 Vidarabine therapy of neonatal herpes simplex virus infection. Pediatrics, 66, 495–501. Whitley, R. J., Nahmias, A. J., Visintine, A. M., Fleming, C. L., & Alford, C. A. (1980 The natural history of herpes simplex virus infection of mother and newborn. Pediatrics, 66, 489–494. Workowski, K. A., & Berman, S. (2010 Sexually transmitted diseases treatment guidelines. MMWR - Recommendations and Reports, 59, 1–110 [doi: rr5912a1 [pii]].

Hospitalization cost per case of neonatal herpes simplex virus infection from claims data.

The purpose of this study was to estimate the average excess inpatient cost of neonatal herpes simplex virus (NHSV) infection from 2005 to 2009 insura...
138KB Sizes 0 Downloads 7 Views