Journal of Clinical Neuroscience 21 (2014) 1874–1880
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Neurologic disorders, in-hospital deaths, and years of potential life lost in the USA, 1988–2011 Benjamin P. Rosenbaum a,b,⇑, Michael L. Kelly a, Varun R. Kshettry a, Robert J. Weil a,c,d a
Department of Neurosurgery, Neurological Institute, Cleveland Clinic, S40, 9500 Euclid Avenue, Cleveland, OH 44195, USA Department of Medical Informatics & Clinical Epidemiology, Oregon Health & Science University, Central Campus, Biomedical Information Communication Center (BICC), Portland, OR, USA c Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center, Cleveland Clinic, Cleveland, OH, USA d Department of Neurosurgery, Geisinger Health System, Danville, PA, USA b
a r t i c l e
i n f o
Article history: Received 2 April 2014 Accepted 11 May 2014
Keywords: Hospital mortality Nationwide Inpatient Sample Neurologic disease burden Premature mortality Public health Traumatic brain injury
a b s t r a c t Premature mortality is a public health concern that can be quantified as years of potential life lost (YPLL). Studying premature mortality can help guide hospital initiatives and resource allocation. We investigated the categories of neurologic and neurosurgical conditions associated with in-hospital deaths that account for the highest YPLL and their trends over time. Using the Nationwide Inpatient Sample (NIS), we calculated YPLL for patients hospitalized in the USA from 1988 to 2011. Hospitalizations were categorized by related neurologic principal diagnoses. An estimated 2,355,673 in-hospital deaths accounted for an estimated 25,598,566 YPLL. The traumatic brain injury (TBI) category accounted for the highest annual mean YPLL at 361,748 (33.9% of total neurologic YPLL). Intracerebral hemorrhage, cerebral ischemia, subarachnoid hemorrhage, and anoxic brain damage completed the group of five diagnoses with the highest YPLL. TBI accounted for 12.1% of all inflation adjusted neurologic hospital charges and 22.4% of inflation adjusted charges among neurologic deaths. The in-hospital mortality rate has been stable or decreasing for all of these diagnoses except TBI, which rose from 5.1% in 1988 to 7.8% in 2011. Using YPLL, we provide a framework to compare the burden of premature in-hospital mortality on patients with neurologic disorders, which may prove useful for informing decisions related to allocation of health resources or research funding. Considering premature mortality alone, increased efforts should be focused on TBI, particularly in and related to the hospital setting. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Premature mortality, as defined by years of potential life lost (YPLL) and life expectancy, is a significant public health concern in the USA [1–3]. Diagnoses that account for the highest YPLL may be an area for increased public health attention, resource allocation, or quality improvement. A focused evaluation of YPLL among neurologic disorders has not been undertaken to our knowledge. YPLL has been specifically studied in several discrete populations including unintentional child and adolescent injuries [4,5], some traumatic populations [6], and alcohol-related deaths [7]. Unintentional injuries are the leading cause of deaths in young people and contributes significantly to overall YPLL calculations [4,5]. Traumatic brain injury (TBI) and spinal cord injury as a result ⇑ Corresponding author. Tel.: +1 216 444 5539. E-mail address: [email protected] (B.P. Rosenbaum). http://dx.doi.org/10.1016/j.jocn.2014.05.006 0967-5868/Ó 2014 Elsevier Ltd. All rights reserved.
of gunshot injury are estimated to shorten lifespan by 3.1 days in the USA [6]. Alcohol also contributes significantly to premature mortality, with an estimated 1,288,700 YPLL in the USA in 2005 [7]. Most recently, the Global Burden of Disease Study 2010 (GBD 2010) analyzed causes of death and premature mortality including YPLL in 1990 and 2010 on a large scale in multiple countries [8,9]. A variety of related analyses, including the USA Burden of Disease Collaborators, have been published more recently [10–13]. The GBD 2010 data demonstrated stroke as an important contributor to YPLL with 2,250,400 YPLL in 1990 and 1,945,300 YPLL in 2010 in the USA. According to the National Center for Health Statistics (NCHS), in 2000 and 2010, approximately one-third of all USA deaths occurred in short-stay, general hospitals [13]. Given these facts, we evaluated the scale of premature mortality limited to USA inhospital mortality of neurologic diagnoses using the YPLL framework. We sought to identify the neurologic diseases responsible for the greatest inpatient YPLL over a longitudinal timespan to
1875
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880 Table 1 Neurologic principal diagnosis disease categories Category
Number of diagnoses*
Anoxic brain damage Brain tumor Cerebral edema and brain compression Cerebral ischemia CNS infection Cognitive deficit Coma - unspecified Congenital Degenerative spine Demyelinating Epidural hematoma Epilepsy Headache Intracerebral hemorrhage Neurologically degenerative Neuropathy and myopathy Paralysis Parkinson’s disease Spine trauma Spine tumor Subarachnoid hemorrhage Subdural hemorrhage (non-traumatic) Traumatic brain injury Unruptured aneurysm CCS categories with excluded diagnoses
1 15 2 46 124 42 10 29 64 5 1 31 66 2 63 96 40 2 80 2 1 1 494 1 235
CCS categories**
ICD-9-CM diagnosis codes from partial CCS categories
85p 35p, 95p 95p 109p, 110, 111p, 112 76, 77, 78, 95p 95p, 653 85p, 95p 82p, 216 205p 80, 95p 109p 83 84 109p 81p, 95p 81p, 95p, 205p 82p 79, 95p 205p, 227 35p 109p 109p 228p, 233 111p 11, 35p, 95p, 113, 228p
348.1 191.XX, 192, 192.1, 348.0, V10.85, V12.41 348.4, 348.5 346.XX, 433.XX, 434.XX, 436, 437.XX (except 437.3) V12.42 331.83, 799.51, 799.52, 799.55, 799.59 348.3X, 348.82, 349.82, 780.XX 343.XX All except 721.7, 723.4, 723.5 341.XX 432.0
431, 432.9 330.XX, 331.XX, 333.XX-336.XX, 337.0, 348.2 337.XX, 350.XX - 359.XX (except 353.2, 353.3, 353.4), 723.4, 723.5, 781.7 All except 343.XX 332.1 721.7 192.2, 192.3 430 432.1 800.XX, 801.XX, 802.0, 802.1, 803.XX, 804.XX, 905.0 437.3
CCS = clinical classification software, CNS = central nervous system, ICD-9-CM = International Classification of Diseases, Ninth Revision, Clinical Modification. * The number of diagnoses within a category represents coding nomenclature and does not necessarily indicate an increased likelihood of admissions being assigned to the category. ** p = only part of the CCS category diseases included; explained in next column.
understand the hospital contribution to YPLL in patients afflicted with neurologic conditions. 2. Methods 2.1. Study design We examined all in-patient hospitalizations from 1988 to 2011 using hospital discharge data from the Nationwide Inpatient Sample (NIS), Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality (Rockville, MD, USA). The NIS is the largest publicly available, all-payer inpatient care database of non-federal hospitals in the USA. The NIS contains data from approximately 5 to 8 million hospital stays each year and represents an approximately 20% stratified sample of USA community (non-federal) hospitals. NIS data are represented by diagnosis and procedure codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). Hospitalizations were categorized by principal diagnoses using 24 custom category
definitions. Categories were initially identified by selecting the 24 neurologic-related HCUP Clinical Classification Software (CCS) categories. Because CCS categories are not completely aligned with clinical neurologic diseases, we re-classified some of the diagnoses in CCS categories according to clinical disease states (Table 1). Some diagnoses contained within the CCS categories were excluded (for example, pain or sleep disorders, or diagnoses lacking specificity; Table 1). Constituent ICD-9-CM codes in each CCS category are readily available (see http://www.hcup-us.ahrq.gov/ toolssoftware/ccs/ccs.jsp). Hospitalizations with a missing or invalid age were excluded (0.08%; 131,108 of the total NIS sample of 171,998,419). NIS discharge weight was used to extrapolate to estimates of annual USA volume (NIS variable DISCWT). Corrected discharge weight data for years 1988 to 1997, as suggested and published by HCUP, were used (see http://www.hcup-us.ahrq.gov/db/ nation/nis/trendwghts.jsp). Total hospitalization charges were corrected for inflation to the 2011 value using the USA Bureau of Labor Statistics Consumer Price Index Inflation Calculator [14].
Table 2 Ten neurologic principal diagnosis categories causing the highest years of potential life lost as a result of in-hospital death between 1988 and 2011 Principal diagnosis disease category
Traumatic brain injury Intracerebral hemorrhage Cerebral ischemia Subarachnoid hemorrhage Anoxic brain damage CNS infection Brain tumor Epilepsy Subdural hemorrhage (non-traumatic) Coma - unspecified
Rank
1 2 3 4 5 6 7 8 9 10
Annual mean YPLL ± SD*
% of total neurologic YPLL
Annual mean deaths*
Mortality rate
Mean inflation adjusted total charges among deaths
Inflation adjusted total charges of deaths per YPLL**
361,748 ± 39,569 184,819 ± 14,582 151,034 ± 14,038 103,229 ± 10,722 47,158 ± 8,888 44,693 ± 12,958 42,425 ± 9,821 35,457 ± 7,634 16,332 ± 4,527 15,444 ± 6,081
33.9% 17.3% 14.2% 9.7% 4.4% 4.2% 4.0% 3.3% 1.5% 1.4%
15,051 21,116 35,365 6,317 2,533 3,163 1,963 2,595 2,354 1,697
7.5% 30.4% 4.3% 25.5% 52.1% 2.0% 5.5% 1.0% 12.5% 4.3%
$ $ $ $ $ $ $ $ $ $
$ $ $ $ $ $ $ $ $ $
Monetary values are in USD. CNS = central nervous system, SD = standard deviation, YPLL = years of potential life lost. * National values are estimated using Nationwide Inpatient Sample discharge weight. ** In hospitalizations where the patient died, this represents the total inflation adjusted in-hospital charges per total YPLL.
62,252 31,274 33,768 61,851 41,734 58,522 52,370 51,693 50,772 42,039
2,590 3,538 7,741 3,757 2,257 4,150 2,429 3,752 7,256 4,660
1876
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Fig. 1. (A) Disease incidence for all hospitalizations by diagnosis category. (B) In-hospital mortality rate by diagnosis category. YPLL = years of potential life lost. (This figure is available in colour at http://www.sciencedirect.com.)
2.2. YPLL YPLL is a measure of premature mortality and calculates the number of years a person would have lived had he or she not died prematurely. YPLL was calculated for each individual hospitalization by determining the age of patients with in-hospital mortality
and subtracting from a reference age. The reference age was determined for each hospitalization by year and sex according to the National Vital Statistics Reports: Deaths: Final Data for 2010 and Deaths: Preliminary Data for 2011 life expectancy [15,16]. If the patient’s age at the time of death was greater than the reference age, YPLL was set to zero. For example, the life expectancy
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
1877
Fig. 2. Annual estimated years of potential life lost (YPLL) by principal diagnosis category. Estimate created using Nationwide Inpatient Sample (NIS) variable discharge weight (DISCWT). (This figure is available in colour at http://www.sciencedirect.com.)
(reference age) for a woman in 2010 was 81 years. In 2010, a woman who died at age 64 years would have 17 YPLL; a woman who died at age 83 years would have 0 YPLL. In addition, given the longitudinal nature of the data, the variable reference age accounts for changing population mean age and expected age of death between 1988 and 2011. Other methods exist to calculate YPLL, such as a single global reference age [17]. Given the long time span in the study, an annual, sex-based reference age was chosen given the rise in USA life expectancy over time. No age-weighting or discounting methods were employed.
3. Results We reviewed NIS data on 11,967,983 (7.0% of overall NIS) hospitalizations between 1988 and 2011 with valid age data belonging to the neurologic disease categories (Table 1). Using discharge weighting over the 24 year period, there were an estimated national 25,598,566 YPLL for an estimated 2,355,673 in-hospital mortalities (mean 10.9 YPLL per mortality). Neurologic diagnoses accounted for 12.4% of the overall estimated in-hospital YPLL in the USA. Table 2 illustrates the neurologic diagnosis categories with the 10 highest YPLL over the study period. The top five diagnostic categories were TBI, intracerebral hemorrhage (ICH), cerebral ischemia, subarachnoid hemorrhage (SAH), and anoxic brain damage. These categories account for 79.5% of the total neurologic YPLL between 1988 and 2011. The in-hospital incidence of each diagnosis category appears relatively stable over time (Fig. 1A). Among the top five diagnosis categories, the in-hospital mortality rate has been increasing for TBI, remaining relatively stable for anoxic brain damage, and decreasing for cerebral ischemia, ICH, and SAH (Fig. 1B). In addition, the annual YPLL for the top five diagnosis categories appears to be increasing for ICH, decreasing for TBI and
cerebral ischemia, and relatively stable for SAH and anoxic brain damage (Fig. 2). The TBI diagnosis category accounted for the highest annual mean YPLL (361,748) representing 33.9% of total neurologic YPLL (Table 2). Anoxic brain damage had the highest in-hospital mortality rate (52.1%), but resulted in the fifth highest annual mean YPLL (47,158). For all in-hospital mortality, the inflation adjusted total in-hospital charges (in USD) per YPLL was highest for cerebral ischemia ($7741 per YPLL) and lowest for anoxic brain damage ($2257 per YPLL). TBI accounted for 12.1% of all inflation adjusted neurologic hospital charges and 22.4% of inflation adjusted charges among neurologic deaths. For the top five diagnosis categories, overall and YPLL-specific inflation adjusted total in-hospital charges increased over the study period (Fig. 3). Table 3 lists the mean length of stay, age, and in-hospital death along with the proportion of male sex for the principal diagnosis categories accounting for the 10 highest YPLL. Of those categories, the highest mean length of stay of in-hospital deaths was 15.4 days for brain tumors. The highest and lowest mean age of in-hospital deaths was 77.3 years for cerebral ischemia and 54.4 years for TBI. The highest mean YPLL per in-hospital death was 24.0 for TBI and lowest at 4.3 for cerebral ischemia diagnosis categories. Males accounted for 65% of TBI hospitalizations, but only 37% of SAH hospitalizations.
4. Discussion Our results demonstrate the principal diagnosis categories associated with the highest inpatient YPLL for neurologic disorders between 1988 and 2011. TBI represented 33.9% of the total YPLL due to in-hospital deaths from all neurologic conditions. TBI was less common than cerebral ischemia and had a significantly lower mortality rate than ICH and SAH; however, because it had the lowest mean age and greatest YPLL per in-hospital death, TBI accounts
1878
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Fig. 3. Annual inflation adjusted total in-hospital charges (in USD) for (A) all neurologic hospitalizations and (B) per years of potential life lost (YPLL) for in-hospital mortality. (This figure is available in colour at http://www.sciencedirect.com.)
for the greatest overall YPLL. The estimated number of annual deaths from TBI was lower in our study (15,051) than that estimated by the Centers for Disease Control and Prevention (50,000) likely due to the fact the latter includes out of hospital deaths [18]. An estimated half to two-thirds of TBI related mortalities occur
before hospital admission [19,20]; and, for survivors of TBI, the risk of death is increased for at least several years after the injury [21– 23]. Although the overall number of deaths in the USA due to TBI has reportedly decreased in the last three decades, several studies show that the mortality rate for hospitalized TBI patients has been
1879
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Table 3 Length of stay, age, and in-hospital death along with male incidence for the 10 principal diagnosis categories causing the highest years of potential life lost during neurologic Nationwide Inpatient Sample hospitalizations between 1988 and 2011 Principal diagnosis disease category
Overall mean length of stay (days)
Mean length of stay of in-hospital deaths (days)
Overall mean age (years)
Mean age of in-hospital deaths (years)
Mean YPLL per in-hospital death
Male sex
Male sex (in-hospital deaths)
Traumatic brain injury Intracerebral hemorrhage Cerebral ischemia Subarachnoid hemorrhage Anoxic brain damage CNS infection Brain tumor Epilepsy Subdural (non-traumatic) Coma – unspecified
6.5 8.9 5.6 13.0 16.3 5.0 8.1 4.1 8.4 6.2
5.9 4.9 9.4 6.5 9.3 11.4 15.4 11.7 7.5 10.1
43.1 69.0 72.1 57.1 55.9 47.5 50.2 41.3 70.1 66.2
54.4 70.9 77.3 62.5 59.7 65.6 55.5 65.5 73.8 71.6
24.0 8.8 4.3 16.3 18.6 14.1 21.6 13.7 6.9 9.1
65% 49% 46% 37% 55% 41% 55% 51% 62% 45%
67% 48% 44% 35% 53% 50% 56% 49% 56% 50%
CNS = central nervous system, YPLL = years of potential life lost.
stable or increasing [24–27]. Using Centers for Disease Control and Prevention TBI data from 1997–2007, Coronado et al. illustrated decreasing rates of death upon arrival and decreasing mortality rates for patients who arrived in the emergency department but stable mortality rates for in-patient deaths [25]. Our study and Farhad et al. demonstrate that the in-hospital mortality rate for TBI increased during the same time period [28]. Conversely, our study found a steady decrease of in-hospital mortality rates for ICH, SAH, and cerebral ischemia (Fig. 1B). This finding has been corroborated by other studies [12,29–31]. The fact that TBI results in the greatest public health burden in terms of YPLL due to in-hospital death while other neurologic disorders have shown improvements in mortality and YPLL over time suggests that TBI represents a continued target for clinical and resource allocation [32,33]. Increasing TBI in-hospital mortality may also be a result of improved pre-hospital care, but it has been suggested that in-hospital care has trailed the advances made in the triage and transport of TBI patients prior to hospitalization [28,34–36]. The National Institutes of Health publishes data on annual categorical disease spending [37]. In 2011, TBI received $81 million, whereas stroke received $317 million, brain cancer $280 million, neurodegenerative diseases $1,622 million, and neuropathy $146 million. These funding distributions have remained consistent over time and have been similarly demonstrated by others [38–40]. While reducing mortality and minimizing YPLL is an important goal, we recognize a variety of other goals (such as quality of life) in treating an array of neurologic (and other) disorders. Although the direct effects of charges on in-hospital mortality are unclear, the difference in charges between cerebral ischemia ($7741 per YPLL) and TBI ($2590 per YPLL) may also demonstrate a disparity in resource allocation (Table 2). Finally, comparison of USA and international data suggests that health outcomes for diseases of the nervous system, in terms of YPLL, trail those of other developed countries [10,13,41–43]. 5. Limitations Traditional limits of population-based, retrospective data sets relying on codified billing data exist. In some disorder classes, the underlying etiology for the condition, such as anoxic brain injury, are not specified. The annual variation in incidence of each principal diagnosis may be due, in part, to documentation, coding, and/or billing practices or incentives that have changed over time. However, our data provide a large, longitudinal, and representative sample of disease incidence across the USA population. In addition, the current analysis does not consider other epidemiologic markers of disease severity and impact on public health, such as healthy
life expectancy, years lived with disability, disability-adjusted life years (calculated as the sum of YPLL and years lived with disability), or potential gains in life expectancy [10,13,41,44,45]. We focused on YPLL given the overall perspective provided when comparing multiple disease states over time. For example, using YPLL, the state of Florida county health agencies planned programs to reduce premature mortality [46]. Our work and the YPLL framework can help shape similar program development. Additional factors, such as medical comorbidities and extra-cranial injuries, may also influence mortality rates in each disease state. However, an analysis of these factors remains outside the scope of the present study. This work focused solely on in-hospital mortality, not overall long-term mortality. The in-hospital limitation is a result of utilizing the NIS database as the data source. Despite the limitation, the NIS is a large, robust data source that reflects the trends in USA hospitalizations over a long time period. In addition, according to the NCHS, in 2000 and 2010, approximately one-third of all USA deaths occurred in short-stay, general hospitals [13]. The actual national hospital-related mortality rate is likely higher if dying patients, fully treated in hospital but who are transferred out-ofhospital, are considered. The YPLL analysis of patients who survive to discharge is another group entirely, upon which separate conclusions may be drawn.
6. Conclusions YPLL provides a method to compare the effects of various neurologic disorders on public health and longevity. TBI, followed by ICH, cerebral ischemia, SAH, and anoxic brain damage account for the highest in-hospital YPLL. The in-hospital mortality rate has been increasing for TBI, remaining relatively stable for anoxic brain damage, and decreasing for cerebral ischemia, ICH, and SAH. The data presented highlighting neurologic principal diagnoses and their associated incidence, in-hospital mortality rates, lengths of stay, mean ages, and total in-hospital charges may help identify targets for increased investigation or resource allocation. Considering premature mortality alone, increased efforts should be focused on TBI, particularly in and related to the hospital setting.
Conflicts of Interest/Disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication.
1880
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Acknowledgements R.J.W. was supported in part by the Melvin Burkhardt Chair in Neurosurgical Oncology, and by the Karen Colina Wilson Research Endowment within the Brain Tumor and Neuro-oncology Center at the Cleveland Clinic. The funders of this philanthropic support had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. References [1] Centre for Disease Control (CDC). Premature mortality in the United States: public health issues in the use of years of potential life lost. MMWR Morb Mortal Wkly Rep 1986;35:1S–11S. [2] Shkolnikov VM, Andreev EM, Zhang Z, et al. Losses of expected lifetime in the United States and other developed countries: methods and empirical analyses. Demography 2011;48:211–39. [3] Wang H, Schumacher AE, Levitz CE, et al. Left behind: widening disparities for males and females in US county life expectancy, 1985–2010. Popul Health Metr 2013;11:8. [4] Centre for Disease Control (CDC). Years of potential life lost from unintentional injuries among persons aged 0-19 years – United States, 2000–2009. MMWR Morb Mortal Wkly Rep 2012;61:830–3. [5] Borse NN, Rudd RA, Dellinger AM, et al. Years of potential life lost from unintentional child and adolescent injuries–United States, 2000–2009. J Safety Res 2013;45:127–31. [6] Richmond TS, Lemaire J. Years of life lost because of gunshot injury to the brain and spinal cord. Am J Phys Med Rehabil 2008;87:609–15 [quiz 15–8]. [7] Shield KD, Gmel G, Kehoe-Chan T, et al. Mortality and potential years of life lost attributable to alcohol consumption by race and sex in the United States in 2005. PLoS One 2013;8:e51923. [8] Murray CJ, Ezzati M, Flaxman AD, et al. GBD 2010: design, definitions, and metrics. Lancet 2012;380:2063–6. [9] Murray CJ, Ezzati M, Flaxman AD, et al. GBD 2010: a multi-investigator collaboration for global comparative descriptive epidemiology. Lancet 2012;380:2055–8. [10] Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2197–223. [11] Lozano R, Naghavi M, Foreman K, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2095–128. [12] Rincon F, Rossenwasser RH, Dumont A. The epidemiology of admissions of nontraumatic subarachnoid hemorrhage in the United States. Neurosurgery 2013;73:217–22 [discussion 212–3]. [13] Murray CJ, Abraham J, Ali MK, et al. The state of US Health, 1990–2010: burden of diseases, injuries, and risk factors. JAMA 2013;310:591–608. [14] Statistics USBoL. Inflation calculator: bureau of labor statistics; 2012. Available from: http://www.bls.gov/data/inflation_calculator.htm. [cited 01 December 2012]. [15] National Vital Statistics System. In: Hoyert D, Xu J, editors. Deaths: preliminary data for 2011. National Center for Health Statistics; 2012. p. 52. [16] National Vital Statistics System. In: Murphy S, Xu J, Kochanek K, editors. Deaths: final data for 2010. National Center for Health Statistics; 2013. p. 168. [17] Gardner JW, Sanborn JS. Years of potential life lost (YPLL)–what does it measure? Epidemiology 1990;1:322–9. [18] Coronado VG, Faul M, Wald MM, et al. Traumatic brain injury in the United States: emergency department visits, hospitalizations, and deaths, 2002– 2006. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; 2010. [19] Kraus JF, Black MA, Hessol N, et al. The incidence of acute brain injury and serious impairment in a defined population. Am J Epidemiol 1984;119: 186–201.
[20] Bruns J Jr, Hauser WA. The epidemiology of traumatic brain injury: a review. Epilepsia 2003;44:2–10. [21] McMillan TM, Teasdale GM. Death rate is increased for at least 7 years after head injury: a prospective study. Brain 2007;130:2520–7. [22] Rutherford GW, Wlodarczyk RC. Distant sequelae of traumatic brain injury: premature mortality and intracranial neoplasms. J Head Trauma Rehabil 2009;24:468–74. [23] Pentland B, Hutton LS, Jones PA. Late mortality after head injury. J Neurol Neurosurg Psychiatry 2005;76:395–400. [24] Adekoya N, Thurman DJ, White DD, et al. Surveillance for traumatic brain injury deaths–United States, 1989–1998. MMWR Surveill Summ 2002;51:1–14. [25] Coronado VG, Xu L, Basavaraju SV, et al. Surveillance for traumatic brain injury-related deaths–United States, 1997–2007. MMWR Surveill Summ 2011;60:1–32. [26] Sosin DM, Sniezek JE, Waxweiler RJ. Trends in death associated with traumatic brain injury, 1979 through 1992. Success and failure. JAMA 1995;273:1778–80. [27] Stein SC, Georgoff P, Meghan S, et al. 150 years of treating severe traumatic brain injury: a systematic review of progress in mortality. J Neurotrauma 2010;27:1343–53. [28] Farhad K, Khan HM, Ji AB, et al. Trends in outcomes and hospitalization costs for traumatic brain injury in adult patients in the United States. J Neurotrauma 2013;30:84–90. [29] Rincon F, Mayer SA. The epidemiology of intracerebral hemorrhage in the United States from 1979 to 2008. Neurocrit Care 2013;19:95–102. [30] Adeoye O, Ringer A, Hornung R, et al. Trends in surgical management and mortality of intracerebral hemorrhage in the United States before and after the STICH trial. Neurocrit Care 2010;13:82–6. [31] Towfighi A, Ovbiagele B, Saver JL. Therapeutic milestone: stroke declines from the second to the third leading organ- and disease-specific cause of death in the United States. Stroke 2010;41:499–503. [32] The Lancet Neurology. A rally for traumatic brain injury research. Lancet Neurol 2013; 12:1127. [33] Whitmore RG, Thawani JP, Grady MS, et al. Is aggressive treatment of traumatic brain injury cost-effective? J Neurosurg 2012;116:1106–13. [34] Bulger EM, Nathens AB, Rivara FP, et al. National variability in out-of-hospital treatment after traumatic injury. Ann Emerg Med 2007;49:293–301. [35] MacKenzie EJ, Rivara FP, Jurkovich GJ, et al. A national evaluation of the effect of trauma-center care on mortality. N Engl J Med 2006;354:366–78. [36] Tiesman H, Young T, Torner JC, et al. Effects of a rural trauma system on traumatic brain injuries. J Neurotrauma 2007;24:1189–97. [37] National Institutes of Health. Estimates of funding for various research, condition, and disease categories (RCDC); 2013. Available from: http:// report.nih.gov/categorical_spending.aspx. [cited 16 January 2014]. [38] Gillum LA, Gouveia C, Dorsey ER, et al. NIH disease funding levels and burden of disease. PLoS One 2011;6:e16837. [39] Meador KJ, French J, Loring DW, et al. Disparities in NIH funding for epilepsy research. Neurology 2011;77:1305–7. [40] Young WB, Shapiro RE. Disparities in NIH funding for epilepsy research. Neurology 2012;78:292–3. [41] Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2163–96. [42] Moses H 3rd, Matheson DH, Dorsey ER, et al. The anatomy of health care in the United States. JAMA 2013;310:1947–63. [43] IOM. US Health in International Perspective: Shorter Lives, Poorer Health; 2013. Available from: http://www.iom.edu/Reports/2013/US-Health-in-InternationalPerspective-Shorter-Lives-Poorer-Health.aspx. [cited 7 February 2014]. [44] Lai D, Hardy RJ. Potential gains in life expectancy or years of potential life lost: impact of competing risks of death. Int J Epidemiol 1999;28:894–8. [45] Salomon JA, Wang H, Freeman MK, et al. Healthy life expectancy for 187 countries, 1990–2010: a systematic analysis for the Global Burden Disease Study 2010. Lancet 2012;380:2144–62. [46] McDonnell S, Vossberg K, Hopkins RS, et al. Using YPLL in health planning. Public Health Rep 1998;113:55–61.
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Neurologic disorders, in-hospital deaths, and years of potential life lost in the USA, 1988–2011 Benjamin P. Rosenbaum a,b,⇑, Michael L. Kelly a, Varun R. Kshettry a, Robert J. Weil a,c,d a
Department of Neurosurgery, Neurological Institute, Cleveland Clinic, S40, 9500 Euclid Avenue, Cleveland, OH 44195, USA Department of Medical Informatics & Clinical Epidemiology, Oregon Health & Science University, Central Campus, Biomedical Information Communication Center (BICC), Portland, OR, USA c Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center, Cleveland Clinic, Cleveland, OH, USA d Department of Neurosurgery, Geisinger Health System, Danville, PA, USA b
a r t i c l e
i n f o
Article history: Received 2 April 2014 Accepted 11 May 2014
Keywords: Hospital mortality Nationwide Inpatient Sample Neurologic disease burden Premature mortality Public health Traumatic brain injury
a b s t r a c t Premature mortality is a public health concern that can be quantified as years of potential life lost (YPLL). Studying premature mortality can help guide hospital initiatives and resource allocation. We investigated the categories of neurologic and neurosurgical conditions associated with in-hospital deaths that account for the highest YPLL and their trends over time. Using the Nationwide Inpatient Sample (NIS), we calculated YPLL for patients hospitalized in the USA from 1988 to 2011. Hospitalizations were categorized by related neurologic principal diagnoses. An estimated 2,355,673 in-hospital deaths accounted for an estimated 25,598,566 YPLL. The traumatic brain injury (TBI) category accounted for the highest annual mean YPLL at 361,748 (33.9% of total neurologic YPLL). Intracerebral hemorrhage, cerebral ischemia, subarachnoid hemorrhage, and anoxic brain damage completed the group of five diagnoses with the highest YPLL. TBI accounted for 12.1% of all inflation adjusted neurologic hospital charges and 22.4% of inflation adjusted charges among neurologic deaths. The in-hospital mortality rate has been stable or decreasing for all of these diagnoses except TBI, which rose from 5.1% in 1988 to 7.8% in 2011. Using YPLL, we provide a framework to compare the burden of premature in-hospital mortality on patients with neurologic disorders, which may prove useful for informing decisions related to allocation of health resources or research funding. Considering premature mortality alone, increased efforts should be focused on TBI, particularly in and related to the hospital setting. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Premature mortality, as defined by years of potential life lost (YPLL) and life expectancy, is a significant public health concern in the USA [1–3]. Diagnoses that account for the highest YPLL may be an area for increased public health attention, resource allocation, or quality improvement. A focused evaluation of YPLL among neurologic disorders has not been undertaken to our knowledge. YPLL has been specifically studied in several discrete populations including unintentional child and adolescent injuries [4,5], some traumatic populations [6], and alcohol-related deaths [7]. Unintentional injuries are the leading cause of deaths in young people and contributes significantly to overall YPLL calculations [4,5]. Traumatic brain injury (TBI) and spinal cord injury as a result ⇑ Corresponding author. Tel.: +1 216 444 5539. E-mail address: [email protected] (B.P. Rosenbaum). http://dx.doi.org/10.1016/j.jocn.2014.05.006 0967-5868/Ó 2014 Elsevier Ltd. All rights reserved.
of gunshot injury are estimated to shorten lifespan by 3.1 days in the USA [6]. Alcohol also contributes significantly to premature mortality, with an estimated 1,288,700 YPLL in the USA in 2005 [7]. Most recently, the Global Burden of Disease Study 2010 (GBD 2010) analyzed causes of death and premature mortality including YPLL in 1990 and 2010 on a large scale in multiple countries [8,9]. A variety of related analyses, including the USA Burden of Disease Collaborators, have been published more recently [10–13]. The GBD 2010 data demonstrated stroke as an important contributor to YPLL with 2,250,400 YPLL in 1990 and 1,945,300 YPLL in 2010 in the USA. According to the National Center for Health Statistics (NCHS), in 2000 and 2010, approximately one-third of all USA deaths occurred in short-stay, general hospitals [13]. Given these facts, we evaluated the scale of premature mortality limited to USA inhospital mortality of neurologic diagnoses using the YPLL framework. We sought to identify the neurologic diseases responsible for the greatest inpatient YPLL over a longitudinal timespan to
1875
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880 Table 1 Neurologic principal diagnosis disease categories Category
Number of diagnoses*
Anoxic brain damage Brain tumor Cerebral edema and brain compression Cerebral ischemia CNS infection Cognitive deficit Coma - unspecified Congenital Degenerative spine Demyelinating Epidural hematoma Epilepsy Headache Intracerebral hemorrhage Neurologically degenerative Neuropathy and myopathy Paralysis Parkinson’s disease Spine trauma Spine tumor Subarachnoid hemorrhage Subdural hemorrhage (non-traumatic) Traumatic brain injury Unruptured aneurysm CCS categories with excluded diagnoses
1 15 2 46 124 42 10 29 64 5 1 31 66 2 63 96 40 2 80 2 1 1 494 1 235
CCS categories**
ICD-9-CM diagnosis codes from partial CCS categories
85p 35p, 95p 95p 109p, 110, 111p, 112 76, 77, 78, 95p 95p, 653 85p, 95p 82p, 216 205p 80, 95p 109p 83 84 109p 81p, 95p 81p, 95p, 205p 82p 79, 95p 205p, 227 35p 109p 109p 228p, 233 111p 11, 35p, 95p, 113, 228p
348.1 191.XX, 192, 192.1, 348.0, V10.85, V12.41 348.4, 348.5 346.XX, 433.XX, 434.XX, 436, 437.XX (except 437.3) V12.42 331.83, 799.51, 799.52, 799.55, 799.59 348.3X, 348.82, 349.82, 780.XX 343.XX All except 721.7, 723.4, 723.5 341.XX 432.0
431, 432.9 330.XX, 331.XX, 333.XX-336.XX, 337.0, 348.2 337.XX, 350.XX - 359.XX (except 353.2, 353.3, 353.4), 723.4, 723.5, 781.7 All except 343.XX 332.1 721.7 192.2, 192.3 430 432.1 800.XX, 801.XX, 802.0, 802.1, 803.XX, 804.XX, 905.0 437.3
CCS = clinical classification software, CNS = central nervous system, ICD-9-CM = International Classification of Diseases, Ninth Revision, Clinical Modification. * The number of diagnoses within a category represents coding nomenclature and does not necessarily indicate an increased likelihood of admissions being assigned to the category. ** p = only part of the CCS category diseases included; explained in next column.
understand the hospital contribution to YPLL in patients afflicted with neurologic conditions. 2. Methods 2.1. Study design We examined all in-patient hospitalizations from 1988 to 2011 using hospital discharge data from the Nationwide Inpatient Sample (NIS), Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality (Rockville, MD, USA). The NIS is the largest publicly available, all-payer inpatient care database of non-federal hospitals in the USA. The NIS contains data from approximately 5 to 8 million hospital stays each year and represents an approximately 20% stratified sample of USA community (non-federal) hospitals. NIS data are represented by diagnosis and procedure codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). Hospitalizations were categorized by principal diagnoses using 24 custom category
definitions. Categories were initially identified by selecting the 24 neurologic-related HCUP Clinical Classification Software (CCS) categories. Because CCS categories are not completely aligned with clinical neurologic diseases, we re-classified some of the diagnoses in CCS categories according to clinical disease states (Table 1). Some diagnoses contained within the CCS categories were excluded (for example, pain or sleep disorders, or diagnoses lacking specificity; Table 1). Constituent ICD-9-CM codes in each CCS category are readily available (see http://www.hcup-us.ahrq.gov/ toolssoftware/ccs/ccs.jsp). Hospitalizations with a missing or invalid age were excluded (0.08%; 131,108 of the total NIS sample of 171,998,419). NIS discharge weight was used to extrapolate to estimates of annual USA volume (NIS variable DISCWT). Corrected discharge weight data for years 1988 to 1997, as suggested and published by HCUP, were used (see http://www.hcup-us.ahrq.gov/db/ nation/nis/trendwghts.jsp). Total hospitalization charges were corrected for inflation to the 2011 value using the USA Bureau of Labor Statistics Consumer Price Index Inflation Calculator [14].
Table 2 Ten neurologic principal diagnosis categories causing the highest years of potential life lost as a result of in-hospital death between 1988 and 2011 Principal diagnosis disease category
Traumatic brain injury Intracerebral hemorrhage Cerebral ischemia Subarachnoid hemorrhage Anoxic brain damage CNS infection Brain tumor Epilepsy Subdural hemorrhage (non-traumatic) Coma - unspecified
Rank
1 2 3 4 5 6 7 8 9 10
Annual mean YPLL ± SD*
% of total neurologic YPLL
Annual mean deaths*
Mortality rate
Mean inflation adjusted total charges among deaths
Inflation adjusted total charges of deaths per YPLL**
361,748 ± 39,569 184,819 ± 14,582 151,034 ± 14,038 103,229 ± 10,722 47,158 ± 8,888 44,693 ± 12,958 42,425 ± 9,821 35,457 ± 7,634 16,332 ± 4,527 15,444 ± 6,081
33.9% 17.3% 14.2% 9.7% 4.4% 4.2% 4.0% 3.3% 1.5% 1.4%
15,051 21,116 35,365 6,317 2,533 3,163 1,963 2,595 2,354 1,697
7.5% 30.4% 4.3% 25.5% 52.1% 2.0% 5.5% 1.0% 12.5% 4.3%
$ $ $ $ $ $ $ $ $ $
$ $ $ $ $ $ $ $ $ $
Monetary values are in USD. CNS = central nervous system, SD = standard deviation, YPLL = years of potential life lost. * National values are estimated using Nationwide Inpatient Sample discharge weight. ** In hospitalizations where the patient died, this represents the total inflation adjusted in-hospital charges per total YPLL.
62,252 31,274 33,768 61,851 41,734 58,522 52,370 51,693 50,772 42,039
2,590 3,538 7,741 3,757 2,257 4,150 2,429 3,752 7,256 4,660
1876
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Fig. 1. (A) Disease incidence for all hospitalizations by diagnosis category. (B) In-hospital mortality rate by diagnosis category. YPLL = years of potential life lost. (This figure is available in colour at http://www.sciencedirect.com.)
2.2. YPLL YPLL is a measure of premature mortality and calculates the number of years a person would have lived had he or she not died prematurely. YPLL was calculated for each individual hospitalization by determining the age of patients with in-hospital mortality
and subtracting from a reference age. The reference age was determined for each hospitalization by year and sex according to the National Vital Statistics Reports: Deaths: Final Data for 2010 and Deaths: Preliminary Data for 2011 life expectancy [15,16]. If the patient’s age at the time of death was greater than the reference age, YPLL was set to zero. For example, the life expectancy
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
1877
Fig. 2. Annual estimated years of potential life lost (YPLL) by principal diagnosis category. Estimate created using Nationwide Inpatient Sample (NIS) variable discharge weight (DISCWT). (This figure is available in colour at http://www.sciencedirect.com.)
(reference age) for a woman in 2010 was 81 years. In 2010, a woman who died at age 64 years would have 17 YPLL; a woman who died at age 83 years would have 0 YPLL. In addition, given the longitudinal nature of the data, the variable reference age accounts for changing population mean age and expected age of death between 1988 and 2011. Other methods exist to calculate YPLL, such as a single global reference age [17]. Given the long time span in the study, an annual, sex-based reference age was chosen given the rise in USA life expectancy over time. No age-weighting or discounting methods were employed.
3. Results We reviewed NIS data on 11,967,983 (7.0% of overall NIS) hospitalizations between 1988 and 2011 with valid age data belonging to the neurologic disease categories (Table 1). Using discharge weighting over the 24 year period, there were an estimated national 25,598,566 YPLL for an estimated 2,355,673 in-hospital mortalities (mean 10.9 YPLL per mortality). Neurologic diagnoses accounted for 12.4% of the overall estimated in-hospital YPLL in the USA. Table 2 illustrates the neurologic diagnosis categories with the 10 highest YPLL over the study period. The top five diagnostic categories were TBI, intracerebral hemorrhage (ICH), cerebral ischemia, subarachnoid hemorrhage (SAH), and anoxic brain damage. These categories account for 79.5% of the total neurologic YPLL between 1988 and 2011. The in-hospital incidence of each diagnosis category appears relatively stable over time (Fig. 1A). Among the top five diagnosis categories, the in-hospital mortality rate has been increasing for TBI, remaining relatively stable for anoxic brain damage, and decreasing for cerebral ischemia, ICH, and SAH (Fig. 1B). In addition, the annual YPLL for the top five diagnosis categories appears to be increasing for ICH, decreasing for TBI and
cerebral ischemia, and relatively stable for SAH and anoxic brain damage (Fig. 2). The TBI diagnosis category accounted for the highest annual mean YPLL (361,748) representing 33.9% of total neurologic YPLL (Table 2). Anoxic brain damage had the highest in-hospital mortality rate (52.1%), but resulted in the fifth highest annual mean YPLL (47,158). For all in-hospital mortality, the inflation adjusted total in-hospital charges (in USD) per YPLL was highest for cerebral ischemia ($7741 per YPLL) and lowest for anoxic brain damage ($2257 per YPLL). TBI accounted for 12.1% of all inflation adjusted neurologic hospital charges and 22.4% of inflation adjusted charges among neurologic deaths. For the top five diagnosis categories, overall and YPLL-specific inflation adjusted total in-hospital charges increased over the study period (Fig. 3). Table 3 lists the mean length of stay, age, and in-hospital death along with the proportion of male sex for the principal diagnosis categories accounting for the 10 highest YPLL. Of those categories, the highest mean length of stay of in-hospital deaths was 15.4 days for brain tumors. The highest and lowest mean age of in-hospital deaths was 77.3 years for cerebral ischemia and 54.4 years for TBI. The highest mean YPLL per in-hospital death was 24.0 for TBI and lowest at 4.3 for cerebral ischemia diagnosis categories. Males accounted for 65% of TBI hospitalizations, but only 37% of SAH hospitalizations.
4. Discussion Our results demonstrate the principal diagnosis categories associated with the highest inpatient YPLL for neurologic disorders between 1988 and 2011. TBI represented 33.9% of the total YPLL due to in-hospital deaths from all neurologic conditions. TBI was less common than cerebral ischemia and had a significantly lower mortality rate than ICH and SAH; however, because it had the lowest mean age and greatest YPLL per in-hospital death, TBI accounts
1878
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Fig. 3. Annual inflation adjusted total in-hospital charges (in USD) for (A) all neurologic hospitalizations and (B) per years of potential life lost (YPLL) for in-hospital mortality. (This figure is available in colour at http://www.sciencedirect.com.)
for the greatest overall YPLL. The estimated number of annual deaths from TBI was lower in our study (15,051) than that estimated by the Centers for Disease Control and Prevention (50,000) likely due to the fact the latter includes out of hospital deaths [18]. An estimated half to two-thirds of TBI related mortalities occur
before hospital admission [19,20]; and, for survivors of TBI, the risk of death is increased for at least several years after the injury [21– 23]. Although the overall number of deaths in the USA due to TBI has reportedly decreased in the last three decades, several studies show that the mortality rate for hospitalized TBI patients has been
1879
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Table 3 Length of stay, age, and in-hospital death along with male incidence for the 10 principal diagnosis categories causing the highest years of potential life lost during neurologic Nationwide Inpatient Sample hospitalizations between 1988 and 2011 Principal diagnosis disease category
Overall mean length of stay (days)
Mean length of stay of in-hospital deaths (days)
Overall mean age (years)
Mean age of in-hospital deaths (years)
Mean YPLL per in-hospital death
Male sex
Male sex (in-hospital deaths)
Traumatic brain injury Intracerebral hemorrhage Cerebral ischemia Subarachnoid hemorrhage Anoxic brain damage CNS infection Brain tumor Epilepsy Subdural (non-traumatic) Coma – unspecified
6.5 8.9 5.6 13.0 16.3 5.0 8.1 4.1 8.4 6.2
5.9 4.9 9.4 6.5 9.3 11.4 15.4 11.7 7.5 10.1
43.1 69.0 72.1 57.1 55.9 47.5 50.2 41.3 70.1 66.2
54.4 70.9 77.3 62.5 59.7 65.6 55.5 65.5 73.8 71.6
24.0 8.8 4.3 16.3 18.6 14.1 21.6 13.7 6.9 9.1
65% 49% 46% 37% 55% 41% 55% 51% 62% 45%
67% 48% 44% 35% 53% 50% 56% 49% 56% 50%
CNS = central nervous system, YPLL = years of potential life lost.
stable or increasing [24–27]. Using Centers for Disease Control and Prevention TBI data from 1997–2007, Coronado et al. illustrated decreasing rates of death upon arrival and decreasing mortality rates for patients who arrived in the emergency department but stable mortality rates for in-patient deaths [25]. Our study and Farhad et al. demonstrate that the in-hospital mortality rate for TBI increased during the same time period [28]. Conversely, our study found a steady decrease of in-hospital mortality rates for ICH, SAH, and cerebral ischemia (Fig. 1B). This finding has been corroborated by other studies [12,29–31]. The fact that TBI results in the greatest public health burden in terms of YPLL due to in-hospital death while other neurologic disorders have shown improvements in mortality and YPLL over time suggests that TBI represents a continued target for clinical and resource allocation [32,33]. Increasing TBI in-hospital mortality may also be a result of improved pre-hospital care, but it has been suggested that in-hospital care has trailed the advances made in the triage and transport of TBI patients prior to hospitalization [28,34–36]. The National Institutes of Health publishes data on annual categorical disease spending [37]. In 2011, TBI received $81 million, whereas stroke received $317 million, brain cancer $280 million, neurodegenerative diseases $1,622 million, and neuropathy $146 million. These funding distributions have remained consistent over time and have been similarly demonstrated by others [38–40]. While reducing mortality and minimizing YPLL is an important goal, we recognize a variety of other goals (such as quality of life) in treating an array of neurologic (and other) disorders. Although the direct effects of charges on in-hospital mortality are unclear, the difference in charges between cerebral ischemia ($7741 per YPLL) and TBI ($2590 per YPLL) may also demonstrate a disparity in resource allocation (Table 2). Finally, comparison of USA and international data suggests that health outcomes for diseases of the nervous system, in terms of YPLL, trail those of other developed countries [10,13,41–43]. 5. Limitations Traditional limits of population-based, retrospective data sets relying on codified billing data exist. In some disorder classes, the underlying etiology for the condition, such as anoxic brain injury, are not specified. The annual variation in incidence of each principal diagnosis may be due, in part, to documentation, coding, and/or billing practices or incentives that have changed over time. However, our data provide a large, longitudinal, and representative sample of disease incidence across the USA population. In addition, the current analysis does not consider other epidemiologic markers of disease severity and impact on public health, such as healthy
life expectancy, years lived with disability, disability-adjusted life years (calculated as the sum of YPLL and years lived with disability), or potential gains in life expectancy [10,13,41,44,45]. We focused on YPLL given the overall perspective provided when comparing multiple disease states over time. For example, using YPLL, the state of Florida county health agencies planned programs to reduce premature mortality [46]. Our work and the YPLL framework can help shape similar program development. Additional factors, such as medical comorbidities and extra-cranial injuries, may also influence mortality rates in each disease state. However, an analysis of these factors remains outside the scope of the present study. This work focused solely on in-hospital mortality, not overall long-term mortality. The in-hospital limitation is a result of utilizing the NIS database as the data source. Despite the limitation, the NIS is a large, robust data source that reflects the trends in USA hospitalizations over a long time period. In addition, according to the NCHS, in 2000 and 2010, approximately one-third of all USA deaths occurred in short-stay, general hospitals [13]. The actual national hospital-related mortality rate is likely higher if dying patients, fully treated in hospital but who are transferred out-ofhospital, are considered. The YPLL analysis of patients who survive to discharge is another group entirely, upon which separate conclusions may be drawn.
6. Conclusions YPLL provides a method to compare the effects of various neurologic disorders on public health and longevity. TBI, followed by ICH, cerebral ischemia, SAH, and anoxic brain damage account for the highest in-hospital YPLL. The in-hospital mortality rate has been increasing for TBI, remaining relatively stable for anoxic brain damage, and decreasing for cerebral ischemia, ICH, and SAH. The data presented highlighting neurologic principal diagnoses and their associated incidence, in-hospital mortality rates, lengths of stay, mean ages, and total in-hospital charges may help identify targets for increased investigation or resource allocation. Considering premature mortality alone, increased efforts should be focused on TBI, particularly in and related to the hospital setting.
Conflicts of Interest/Disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication.
1880
B.P. Rosenbaum et al. / Journal of Clinical Neuroscience 21 (2014) 1874–1880
Acknowledgements R.J.W. was supported in part by the Melvin Burkhardt Chair in Neurosurgical Oncology, and by the Karen Colina Wilson Research Endowment within the Brain Tumor and Neuro-oncology Center at the Cleveland Clinic. The funders of this philanthropic support had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. References [1] Centre for Disease Control (CDC). Premature mortality in the United States: public health issues in the use of years of potential life lost. MMWR Morb Mortal Wkly Rep 1986;35:1S–11S. [2] Shkolnikov VM, Andreev EM, Zhang Z, et al. Losses of expected lifetime in the United States and other developed countries: methods and empirical analyses. Demography 2011;48:211–39. [3] Wang H, Schumacher AE, Levitz CE, et al. Left behind: widening disparities for males and females in US county life expectancy, 1985–2010. Popul Health Metr 2013;11:8. [4] Centre for Disease Control (CDC). Years of potential life lost from unintentional injuries among persons aged 0-19 years – United States, 2000–2009. MMWR Morb Mortal Wkly Rep 2012;61:830–3. [5] Borse NN, Rudd RA, Dellinger AM, et al. Years of potential life lost from unintentional child and adolescent injuries–United States, 2000–2009. J Safety Res 2013;45:127–31. [6] Richmond TS, Lemaire J. Years of life lost because of gunshot injury to the brain and spinal cord. Am J Phys Med Rehabil 2008;87:609–15 [quiz 15–8]. [7] Shield KD, Gmel G, Kehoe-Chan T, et al. Mortality and potential years of life lost attributable to alcohol consumption by race and sex in the United States in 2005. PLoS One 2013;8:e51923. [8] Murray CJ, Ezzati M, Flaxman AD, et al. GBD 2010: design, definitions, and metrics. Lancet 2012;380:2063–6. [9] Murray CJ, Ezzati M, Flaxman AD, et al. GBD 2010: a multi-investigator collaboration for global comparative descriptive epidemiology. Lancet 2012;380:2055–8. [10] Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2197–223. [11] Lozano R, Naghavi M, Foreman K, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2095–128. [12] Rincon F, Rossenwasser RH, Dumont A. The epidemiology of admissions of nontraumatic subarachnoid hemorrhage in the United States. Neurosurgery 2013;73:217–22 [discussion 212–3]. [13] Murray CJ, Abraham J, Ali MK, et al. The state of US Health, 1990–2010: burden of diseases, injuries, and risk factors. JAMA 2013;310:591–608. [14] Statistics USBoL. Inflation calculator: bureau of labor statistics; 2012. Available from: http://www.bls.gov/data/inflation_calculator.htm. [cited 01 December 2012]. [15] National Vital Statistics System. In: Hoyert D, Xu J, editors. Deaths: preliminary data for 2011. National Center for Health Statistics; 2012. p. 52. [16] National Vital Statistics System. In: Murphy S, Xu J, Kochanek K, editors. Deaths: final data for 2010. National Center for Health Statistics; 2013. p. 168. [17] Gardner JW, Sanborn JS. Years of potential life lost (YPLL)–what does it measure? Epidemiology 1990;1:322–9. [18] Coronado VG, Faul M, Wald MM, et al. Traumatic brain injury in the United States: emergency department visits, hospitalizations, and deaths, 2002– 2006. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; 2010. [19] Kraus JF, Black MA, Hessol N, et al. The incidence of acute brain injury and serious impairment in a defined population. Am J Epidemiol 1984;119: 186–201.
[20] Bruns J Jr, Hauser WA. The epidemiology of traumatic brain injury: a review. Epilepsia 2003;44:2–10. [21] McMillan TM, Teasdale GM. Death rate is increased for at least 7 years after head injury: a prospective study. Brain 2007;130:2520–7. [22] Rutherford GW, Wlodarczyk RC. Distant sequelae of traumatic brain injury: premature mortality and intracranial neoplasms. J Head Trauma Rehabil 2009;24:468–74. [23] Pentland B, Hutton LS, Jones PA. Late mortality after head injury. J Neurol Neurosurg Psychiatry 2005;76:395–400. [24] Adekoya N, Thurman DJ, White DD, et al. Surveillance for traumatic brain injury deaths–United States, 1989–1998. MMWR Surveill Summ 2002;51:1–14. [25] Coronado VG, Xu L, Basavaraju SV, et al. Surveillance for traumatic brain injury-related deaths–United States, 1997–2007. MMWR Surveill Summ 2011;60:1–32. [26] Sosin DM, Sniezek JE, Waxweiler RJ. Trends in death associated with traumatic brain injury, 1979 through 1992. Success and failure. JAMA 1995;273:1778–80. [27] Stein SC, Georgoff P, Meghan S, et al. 150 years of treating severe traumatic brain injury: a systematic review of progress in mortality. J Neurotrauma 2010;27:1343–53. [28] Farhad K, Khan HM, Ji AB, et al. Trends in outcomes and hospitalization costs for traumatic brain injury in adult patients in the United States. J Neurotrauma 2013;30:84–90. [29] Rincon F, Mayer SA. The epidemiology of intracerebral hemorrhage in the United States from 1979 to 2008. Neurocrit Care 2013;19:95–102. [30] Adeoye O, Ringer A, Hornung R, et al. Trends in surgical management and mortality of intracerebral hemorrhage in the United States before and after the STICH trial. Neurocrit Care 2010;13:82–6. [31] Towfighi A, Ovbiagele B, Saver JL. Therapeutic milestone: stroke declines from the second to the third leading organ- and disease-specific cause of death in the United States. Stroke 2010;41:499–503. [32] The Lancet Neurology. A rally for traumatic brain injury research. Lancet Neurol 2013; 12:1127. [33] Whitmore RG, Thawani JP, Grady MS, et al. Is aggressive treatment of traumatic brain injury cost-effective? J Neurosurg 2012;116:1106–13. [34] Bulger EM, Nathens AB, Rivara FP, et al. National variability in out-of-hospital treatment after traumatic injury. Ann Emerg Med 2007;49:293–301. [35] MacKenzie EJ, Rivara FP, Jurkovich GJ, et al. A national evaluation of the effect of trauma-center care on mortality. N Engl J Med 2006;354:366–78. [36] Tiesman H, Young T, Torner JC, et al. Effects of a rural trauma system on traumatic brain injuries. J Neurotrauma 2007;24:1189–97. [37] National Institutes of Health. Estimates of funding for various research, condition, and disease categories (RCDC); 2013. Available from: http:// report.nih.gov/categorical_spending.aspx. [cited 16 January 2014]. [38] Gillum LA, Gouveia C, Dorsey ER, et al. NIH disease funding levels and burden of disease. PLoS One 2011;6:e16837. [39] Meador KJ, French J, Loring DW, et al. Disparities in NIH funding for epilepsy research. Neurology 2011;77:1305–7. [40] Young WB, Shapiro RE. Disparities in NIH funding for epilepsy research. Neurology 2012;78:292–3. [41] Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2163–96. [42] Moses H 3rd, Matheson DH, Dorsey ER, et al. The anatomy of health care in the United States. JAMA 2013;310:1947–63. [43] IOM. US Health in International Perspective: Shorter Lives, Poorer Health; 2013. Available from: http://www.iom.edu/Reports/2013/US-Health-in-InternationalPerspective-Shorter-Lives-Poorer-Health.aspx. [cited 7 February 2014]. [44] Lai D, Hardy RJ. Potential gains in life expectancy or years of potential life lost: impact of competing risks of death. Int J Epidemiol 1999;28:894–8. [45] Salomon JA, Wang H, Freeman MK, et al. Healthy life expectancy for 187 countries, 1990–2010: a systematic analysis for the Global Burden Disease Study 2010. Lancet 2012;380:2144–62. [46] McDonnell S, Vossberg K, Hopkins RS, et al. Using YPLL in health planning. Public Health Rep 1998;113:55–61.
