American Journal of Emergency Medicine (2012) 30, 293-301

Original Contribution

ED visits for drug-related poisoning in the United States, 2007^?,??

Yuxi Xiang ^a, Weiyan Zhao MD, PhD ^a, Huiyun Xiang MD, MPH, PhD ^a,b,

Gary A. Smith MD, DrPH ^a,b,?

^aCenter for Injury Research and Policy, The Research Institute at Nationwide Children’s Hospital, Columbus, OH 43205, USA

^bThe Ohio State University College of Medicine, Columbus, OH 43210, USA

Received 10 August 2010; revised 27 October 2010; accepted 29 November 2010

Abstract

Background: Fatal drug-related poisoning has been well described. However, death data only show the tip of the iceberg of drug-related poisoning as a Public health problem. Using the 2007 Nationwide Emergency Department Sample, this study described the characteristics of emergency department visits for drug-related poisoning in the United States.

Methods: Any ED visit that had an International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis code of 960-979 was defined as a drug-related poisoning case. Intentionality of poisoning was determined by E-codes. Weighted estimates of ED visits were calculated by patient and hospital characteristics, intentionality of poisoning, and selected drug classes. Population rates by sex, age, urban/rural classification, median Household income in patient’s zip code, and hospital region were calculated.

Results: An estimated 699 123 (95% confidence interval, 666 529-731 717) ED visits for drug-related poisoning occurred in 2007. Children 0 to 5 years old had the highest rate for unintentional poisoning (male, 237 per 100 000; female, 218 per 100 000). The rate of drug-related poisoning in rural areas (684 per 100 000) was 3 times higher than the rates in other areas. Psychotropic agents and analgesics were responsible for 43.7% of all drug-related poisoning. Women 18 to 20 years old had the highest ED visit rate for suicidal poisoning (245 per 100 000). The estimated ED charges were $1 394 051 262, and 41.1% were paid by Medicaid and Medicare.

Conclusion: Antidepressants and analgesics were responsible for nearly 44% of ED visits for drug- related poisoning in the United States. Interventions and future research should target Prescription opioids, rural areas, children 0 to 5 years old for unintentional drug-related poisoning, and female ages 12 to 24 years for suicidal drug-related poisoning.

(C) 2012

^? This study was supported by a grant from the National Center for Injury Prevention and Control, Centers for Disease Control and Prevention (Grant No.

R49/CE001172-01).

^?? The views expressed here are solely the responsibility of the authors and do not necessarily reflect the official views of the Centers for Disease Control and Prevention.

* Corresponding author. Center for Injury Research and Policy, The Research Institute at Nationwide Children’s Hospital, Columbus, OH 43205, USA.

Tel.: +1 614 722 2400; fax: +1 614 722 2448.

E-mail address: [email protected] (G.A. Smith).

0735-6757/$ - see front matter (C) 2012 doi:10.1016/j.ajem.2010.11.031

Introduction

Drug-related poisoning cases have been consistently rising in the United States in the past decade, which can largely be attributed to the increased use of prescription opioid analgesics [1-3]. Statistics from death certificates in 2007 indicated that 27 658 unintentional drug-related poisoning deaths occurred in the United States, which is an increase of roughly 5-fold since 1990 [4]. Drug-related poisoning deaths are currently second only to motor vehicle crash deaths among the leading causes of injury death in the United States [4,5]. Unlike previous epidemics of drug- related poisoning deaths in the United States, current fatal drug-related poisoning have spread to rural areas and are primarily due to prescription drugs rather than illicit drugs [6]. Although there is a considerable amount of literature describing fatal drug-related poisoning, death data only show a fraction of the drug-related poisonings in the United States

[7] Local health agencies are aware of the drug-related poisoning problem but cite insufficient data as one of the major barriers to address drug-related poisoning [8]. Emergency department (ED) visits are a potential data source for monitoring nonfatal drug poisoning. According to the Drug Abuse Warning Network (DAWN) operated by the Substance Abuse and Mental Health Services Administra- tion, the number of ED visits associated with the nonmedical use of over-the-counter and prescription drugs increased from 0.5 million in 2004 to 1 million in 2008 [9]. Meanwhile, the number of ED visits involving illicit drugs was stable at 1 million from 2004 to 2008 [10].

The goal of this study is to describe the epidemiology of ED visits for drug-related poisoning in the United States using the data from the Nationwide Emergency Department Sample (NEDS), one of the health care utilization Project data sets from the Agency for Healthcare Research and Quality. Understanding the epidemiology of drug-related poisoning can help to identify high-risk groups for whom tailored interventions can be developed.

Methods

Data sources

This study examined ED visits for drug poisoning using discharge data from the 2007 NEDS. The NEDS is the largest all-payer ED database that is publicly available in the United States [11]. It selects a 20% stratified probability sample of all US community, nonrehabilitation hospital- based EDs. The 5 hospital characteristics used for stratifi- cation are United States census region, teaching status, ownership, trauma-level designation, and urban-rural loca- tion. The NEDS data enable analyses of ED utilization patterns and yield national estimates of ED visits. The NEDS data set contains more than 100 clinical and nonclinical

variables for each ED visit, including up to 15 diagnoses codes and 4 E-codes per ED discharge. These internationally recognized standard codes provide valuable insights into the reason for ED visits. The 2007 NEDS was released April 2010; thus, it was the most current data available for this study. The data set includes approximately 27 million ED visits from about 970 hospital-based EDs in 27 states, and it generates national estimates pertaining to more than 120 million ED visits.

Definition of drug-related poisoning

Emergency department visits for drug-related poisoning were identified using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9- CM) codes. In our analysis, we searched through all the Diagnosis codes for each ED visit. Any case having the ICD- 9-CM diagnosis code 960-979 (poisoning by drugs, medicinal and biological substances) was defined as drug- related poisoning in this study. ED visits for nonpharma- ceutical substances, such as heroin (code 965.01) and psychodysleptics including marihuana and derivatives (code 969.6), were included in the study.

Any drug-related poisoning case that had more than 1 class of drug recorded in all diagnosis codes (based on first 3 digits of the ICD-9-CM N codes) was defined as a case with multiple drug involvement.

Intentionality of the drug-related poisoning

E-codes were examined for the intentionality of the poisoning. In our study, 94.7% (145 118/153 219) of ED visits for drug-poisoning were assigned with at least 1 valid E-code. Drug-related poisoning cases with any E-code ranging from E850 to E858 (accidental poisoning by drugs, medicinal substances, and biologicals, 40.1%) were considered unintentional. Suicidal cases were defined with E-codes E950.0-E950.5 (suicide and self-inflicted injury by drugs and medicinal substances, 38.7%). All other cases were combined in our “Other” category if they were assigned with E962.0 (assault by drugs and medicinal substances, 0.1%), if they were coded as E980.0-E980.5 (poisoning by drugs and medicinal substances, undetermined whether accidentally or purposely inflicted, 4.1%), if the E-codes were unrelated to poisoning (11.7%), or if no E-code was recorded (5.3%).

Alcohol involvement in the drug-related poisoning

Any drug-related poisoning cases that had the ICD-9-CM diagnosis code 303.0 (Acute alcoholic intoxication), or 305.0 (nondependent alcohol abuse), or 980 (toxic effect of alcohol) were classified as having alcohol involvement in the drug-related poisoning.

Statistical analyses

The 2007 NEDS provides discharge-level sampling weights to allow national estimates of ED visits from all US community, nonrehabilitation hospitals. In this study, data analyses were conducted using SAS 9.1 (SAS Institute, Cary, NC) [12] and SUDAAN 10.0 (Research Triangle Institute, Research Triangle Park, NC) [13] software to account for the complex survey design and sampling procedures of the NEDS. National frequency and percentage estimates (with 95% confidence intervals [CIs]) of ED visits for drug-related poisoning were calculated by sex, age, urban-rural location of patient’s residence, Median household income in patient’s residence zip code, hospital region, disposition from ED, and primary payer. Population rates of ED visits due to drug-related poisoning per 100 000 population were calculated by age group, sex, and hospital region using publicly available 2007 US Census data [14]. We adopted the same age groups (age in years: 0-5, 6-11, 12-17, 18-20, 21-24, 25-29, 30-34, 35-44, 45-

54, 55-64, 65+) that were used in the 2007 DAWN report [9]. Population rates were calculated by patient’s residence and the median household income in a patient’s zip code using Claritas population estimates [15] as the denominator due to a lack of US Census data by these 2 characteristics.

Because approximately 19% of the ED visits (weighted) were missing information about total ED charges (in dollars) in the 2007 NEDS database, we estimated the national total ED charges for drug-related poisoning by using the product of the number of estimated national cases times the average charge per ED visit. This approach is suggested by the Health Care Utilization Project [11].

To provide information about specific types of drugs responsible for poisoning, we provided the national estimates of the percentage of poisoning cases by major drug classes based on ICD-9-CM codes (first 3 digits) of the first-listed diagnosis. If a patient was admitted to the same hospital (State Inpatient Databases [SID] cases), the first-listed diagnosis was the principal diagnosis responsible for the ED visit. For other patients (State Emergency Department Databases [SEDD] cases), the first-listed diagnosis was not necessarily the principal diagnosis. When the proportional distributions of the major classes of the first-listed diagnoses for SID and SEDD cases were compared, there was no significant difference between the 2 groups; therefore, results that combined SID and SEDD cases are presented in this article. In addition, half of the drug-related poisoning cases had either ICD-9-CM code 965 (poisoning by analgesics, antipyretics, and antirheumatics) or ICD-9-CM code 969 (poisoning by psychotropic agents) as the first-listed diagnosis. Therefore, national estimates and intentionality are presented for these 2 classes of drugs.

Ethical consideration

This study used de-identified publicly available data; therefore, this study was considered institutional review

board exempt by the Institutional Review Board of the Research Institute at Nationwide Children’s Hospital.

Results

Demographics and overall characteristics

According to the drug-related poisoning definition in this study, the NEDS data set included 153 219 drug- related poisoning ED visits, representing 699 123 (95% CI, 666 529-731 717) visits nationwide in 2007. Of these cases,

79.2 % had one drug class involved, 14.3 % had 2 classes of drugs involved, 4.7 % had 3 classes of drugs involved, and 1.8 % had 4 or more classes of drugs involved. Alcohol was involved in 90 886 of these drug-related poisoning cases (13.0%).

More than half of drug-related poisoning cases were among women, with 396 353 estimated visits (56.7%)

nationwide, compared with 302 682 visits by men (43.3%; Table 1). Females had a significantly higher rate of poisoning than males after age of 11 years (Fig. 1). Patients aged 35 to 44 and 45 to 54 years had the highest

number of ED visits of 121 989 (17.5%) and 112 589

(16.1%), respectively, followed by 64 878 visits (9.3%) for the 12- to 17-year age group (Table 1). Young children 0 to 5 years old had the highest rate of ED visits for unintentional drug-related poisoning among all age groups (Fig. 2). Females had much higher rates of drug poisoning related to suicidal intent (112 per 100 000 population) than males (66 per 100 000), especially among 12 to 20 years

old (Fig. 3).

As shown in Table 1, patients most commonly resided in large metropolitan areas (47.0 %), followed by small metropolitan areas (22.7%). Fewer patients were from rural areas (the nonmetropolitan/micropolitan areas), representing 19.1% of the nationwide totals. However, rural areas had the highest rate of drug-related poisoning (684 per 100 000). Compared with other categories, patients in areas with low-median household income ($1- $38 999) had the highest percentage (29.7%) and the highest rate (263 per 100 000) of drug-related poisoning ED visits.

ED disposition and charges

As presented in Table 2, routine discharge (43.0%; 95% CI, 41.7%-44.2%) and admission into the same hospital (41.2%; 95% CI, 40.0%-42.4%) accounted for most dispositions of patients from the ED. Total ED charges for drug-related poisoning ED visits averaged $1994 per visit. For the estimated 699 123 visits in 2007, total ED charges were $1 394 051 262 (95% CI, $1 333 227 561-$1 455 574

086). Payments for more than 40% of the drug-related poisoning ED visits came from Medicaid (23.6%) and Medicare (17.5%). Private insurance (including health

Table 1 Demographics of US ED visits for drug-related poisoning, 2007

Sample size (n) in NEDS		National estimate	Weighted % (95% CI)		Rate (per 100 000)
Total	153 219	699 123	100		232
Sex a
Male	66 354	302 682	43.3	(42.8-43.8)	204
Female	86 848	396 353	56.7	(56.2-57.2)	259
Unknown	b	b	b		c
Age (y) a
0-5	13 588	63 060	9.0	(8.3-9.8)	255
6-11	1433	6614	0.9	(0.8-1.0)	28
12-17	14 110	64 878	9.3	(8.8-9.8)	256
18-20	10 050	45 799	6.6	(6.4-8.8)	363
21-24	11 815	53 530	7.7	(7.5-7.9)	317
25-29	13 813	62 145	8.9	(8.7-9.1)	295
30-34	11 616	52 648	7.5	(7.3-7.7)	270
35-44	26 811	121 989	17.5	(17.0-17.9)	283
45-54	24 627	112 589	16.1	(15.7-16.5)	257
55-64	12 030	54 998	7.9	(7.6-8.1)	168
65+	13 315	60 813	8.7	(8.4-9.0)	161
Unknown	b	b	b		c
Patient’s residence d
Large metropolitan	73 343	328 595	47.0	(44.5-49.5)	203
Small metropolitan	35 712	158 954	22.7	(20.0-25.8)	179
Micropolitan	14 018	71 530	10.2	(8.3-12.5)	233
Not metropolitan/micropolitan	27 705	133 698	19.1	(17.9-20.4)	684
Unknown	1441	6345	0.9	(0.7-1.2)	c
Median household income in patient’s zip code d
$1-$38 999	44 907	207 857	29.7	(27.9-31.7)	263
$39 000-$47 999	40 219	183 758	26.3	(24.8-27.8)	251
$48 000-$62 999	36 757	165 918	23.7	(22.3-25.2)	224
$63 000 or more	26 593	120 359	17.2	(15.6-18.9)	161
Unknown	4743	21 232	3.0	(2.5-3.6)	c
Hospital region a
Northeast	23 703	112 608	16.1	(14.5-17.9)	205
Midwest	35 228	173 275	24.8	(22.8-26.8)	261
South	62 170	266 458	38.1	(35.8-40.5)	241
West	32 118	146 782	21.0	(19.2-22.9)	210
Source: NEDS, 2007. a US census data. b The following statistics were suppressed if (1) sample size was less than or equal to 10 or (2) statistics were based on estimates with a relative standard error greater than 0.30 or with standard error of 0 (population data source). c Population estimates are not available. d Claritas population estimates.

maintenance organization plans) paid for 32.6% of ED visits, and 20.5% of visits were self-paid.

Drugs listed as the first diagnosis

Among drugs listed as the first diagnosis, poisonings by psychotropic agents (ICD-9-CM code 969) were responsible for the highest proportion of ED visits (24.2%; 169 309 cases) (Table 3). The second most common poisoning diagnosis listed as the first diagnosis was poisoning by analgesics, antipyretics, and antirheumatics (ICD-9-CM code 965), accounting for 23.1% (161 510 cases) of ED visits for drug-related poisoning.

Among the estimated 169 309 ED visits due to psychotropic agents, 52.4% were suicidal poisoning, and only 29.5% were unintentional (Table 4). However, 41.1%

(66 381 of 161 510) of poisonings associated with analge- sics, antipyretics, and antirheumatics were suicidal, and 40.1% were unintentional. Although poisonings by psy- chotropic agents had large proportions of cases with suicidal intentions, including antidepressants (61.9%), benzodiazepines (54.1%), and other tranquilizers (59.4%), for some specific classes of analgesics, antipyretics, and antirheumatics, specifically, salicylates (55.%) and aromatic analgesics not elsewhere classified (56.1%), the proportion of ED visits attributed to suicide was nearly as high or

500

450

400

Rate per 100,000 Population

350

300

250

200

150

100

50

0

Females

Males

0-5 6-11 12-17 18-20 21-24 25-29 30-34 35-44 45-54 55-64 65+

Age Group (Years)

300

250

Rate per 100,000 Population

200

150

100

50

0

Females - Intentional Males - Intentional

0-5 6-11 12-17 18-20 21-24 25-29 30-34 35-44 45-54 55-64 65+

Age Group (Years)

Fig. 1 Rate of US ED visits for drug-related poisoning by sex and age group, 2007. Source: NEDS, 2007.

Fig. 3 Rate of US ED visits for drug-related poisoning with suicidal intent by sex and age group, 2007. Source: NEDS, 2007.

higher than for psychotropic agents. This was in contrast to opiates and related narcotics (22.8%), which had lower proportions of cases with suicidal intentionality.

Discussion

With the NEDS data, we estimated that there were 699 123 drug-related poisoning ED visits in the year 2007. The DAWN and National Electronic Injury surveillance System- All Injury Program (NEISS-AIP) are 2 other databases that could be used to monitor and study drug-related poisoning ED visits in the United States [9,16]. They are both probability samples of data from hospital EDs nationally and have been used to study poisoning. However, estimates of the numbers of poisoning cases differ among these 3 databases. DAWN estimated 855 838, and NEISS estimated 880 264 cases in the same year [9,16]. The differences in these estimates could be due to the different definitions of

300

Females - Unintentional Males - Unintentional

250

Rate per 100,000 Population

200

150

100

50

0

0-5 6-11 12-17 18-20 21-24 25-29 30-34 35-44 45-54 55-64 65+

Age Group (Years)

Fig. 2 Rate of US ED visits for unintentional drug-related poisoning by sex and age group, Source: NEDS, 2007.

drug-related poisoning in 3 surveillance systems. For our study, drug-related poisoning was defined as the ingestion, inhalation, absorption through the skin, or injection of a drug that causes harmful results. The DAWN and NEISS-AIP have included additional cases involving a toxin (biological or nonbiological) or other chemical that causes harmful results, which would explain the larger estimated number of poisoning cases using these databases. Variation in the samples for these 3 databases also may have contributed to the observed differences in drug-related poisoning estimates. The NEDS sample includes 970 hospital EDs compared with 220 for the DAWN and 66 for the NEISS-AIP [9,16]. Rates of ED visits for drug-related poisoning due to suicide attempts among females have been previously published for these 2 alternative databases [9,16]. Based on our analysis of the 2007 NEDS, females had a rate of 112/100 000 ED visits for drug-related poisoning compared with 66/100 000 for males. An analysis using 2007 DAWN data indicated that females had a rate of 78/100 000 compared with the male rate of 52/100 000 [9]. Finally, results from the 2007 NEISS data indicated that females had a suicide rate of 80/100 000 compared with the male suicide rate of 50/100 000 [16].

In this study, children 0 to 5 years old had the highest rate of ED visits for unintentional drug-related poisoning (228/ 100 000) compared with other age groups. This finding is consistent with the data from the National Poison Data System (NPDS) of the American Association of Poison control centers, which showed that 51.2% of calls made to participating poison control centers in 2007 were for children 0 to 5 years old [17]. Franklin and Rodgers [18] analyzed 2004 NEISS data and found a high rate of oral drug poisoning (217.9/100 000) among US children 0 to 4 years old. Schillie and colleagues [19] studied ED visits resulting from unintentional medication overdoses among children 18 years or younger using the 2007-2008 NEISS data. In agreement with our findings, Schillie et al [19] found a higher rate of medication overdose among children 0 to 5 years old than among children 6 to 11, 12 to 14, or 15 to

Sample size (n) in NEDS		National estimate	Weighted % (95% CI)
Total	153 219	699 123	100
Disposition of patient from ED
Routine	65 877	300 330	43.0 (41.7-44.2)
Transfer to short-term hospital	6119	28 098	4.0 (3.7-4.4)
Other transfer	13 682	60 983	8.7 (8.0-9.5)
Home health care	534	a	a
Leave against medical advice	1850	8342	1.2 (1.1-1.3)
Admitted into this hospital	62 973	288 120	41.2 (40.0-42.4)
Died in ED	120	577	0.1 (0.07-0.1)
Unknown	2064	10 146	1.5 (1.0-2.1)
Primary payer
Medicare	26 803	122 304	17.5 (17.0-18.0)
Medicaid	35 258	163 080	23.6 (22.4-24.3)
Private including HMO	49 758	228 127	32.6 (31.6-33.7)
Self-pay	31 862	143 351	20.5 (19.5-21.6)
No charge	2301	9142	1.3 (0.8-2.2)
Other	6358	29 775	4.3 (3.8-4.8)
Unknown	799	3344	0.5 (0.4-0.7)
Multiple-drug involvement
1 drug class	121 184	553 701	79.2 (78.6-79.8)
2 drug classes	22 035	100 062	14.3 (13.9-14.7)
3 drug classes	7250	32 959	4.7 (4.6-4.9)
4 drug classes and more	2750	12 400	1.8 (1.7-1.9)
Alcohol involved (yes)	20 005	90 992	13.0 (12.6-13.4)
	Mean	National estimate (95% CI)
Total charges for ED services ($)	1994	1 394 051 262 (1 333 227 561-1 455 574 086)
Source: NEDS, 2007. HMO, health maintenance organization. a The following statistics were suppressed if 1) sample size was less than or equal to 10; or 2) statistics were based on estimates with a relative standard error N0.30 or with standard error=0.

18 years old. However, neither of the studies by Franklin et al

Table 2 Characteristics of US ED visits for drug-related poisoning, 2007

[18] or Schillie et al [19] included persons older than 18 years. Our study included all age groups, and our results indicated that children 0 to 5 years old had a significantly higher rate of ED visits for unintentional drug-related poisoning than any other age group, including adults. Thus, unintentional drug-related poisonings continue to be common among young children despite existence of known successful Prevention strategies, such as child-resistant packaging [20]. These findings underscore the importance of increasing efforts to prevent unintentional drug exposures among US children.

Our finding that 2 classes of drugs (psychotropic agents and analgesics, antipyretics, and antirheumatics) are most commonly involved in ED visits for drug-related poisoning is consistent with the data provided by the DAWN. The growing number of ED cases involving these drugs is associated with a large increase in the sales of methadone, oxycodone, and hydrocodone [21]. These prescription opioids have experienced a 933%, 588%, and 198% increase in sales, respectively, from 1997 to 2005 [1]. According to a

national survey, the percentage of 12^th grade US students who reported having abused prescription opioids in the previous year increased from 3.3% in 1992 to 9.5% in 2004 [22]. In fact, abuse of prescription opioids rose so quickly after 2000 that, in 2007, the annual number of Young people 12 to 17 years old admitted to publicly funded addiction treatment centers for treatment of prescription opioids addiction exceeded the number of admissions for heroin addiction treatment [22]. Kuehn [22] has suggested that a false perception that prescription drugs are safe and the easier access to prescription opioids may be important contributors to many of the poisoning cases caused by prescription opioids. Despite the importance of oxycodone and hydro- codone, ICD-9-CM does not provide unique codes for these drugs as it does for methadone. The International Classification of Diseases, Ninth Revision, Clinical Modifi- cation includes oxycodone and hydrocodone in the category “other opiates and related narcotics.” Thus, it is not possible for us to analyze these drugs separately.

A study by Paulozzi et al [6] suggested that unlike previous epidemics of fatal drug-related poisoning in the

Drug class	ICD-9 code	National estimate	Weighted % (95% CI)
Antibiotics	960	4271	0.6 (0.5-0.7)
Other anti-infectives	961	1792	0.3 (0.2-0.4)
Hormones and synthetic substitutes	962	21 608	3.1 (3.0-3.2)
Primarily systemic agents	963	19 819	2.8 (2.7-3.0)
Agents primarily affecting blood constituents	964	5915	0.9 (0.7-0.9)
Analgesics antipyretics and antirheumatics	965	161 510	23.1 (22.7-23.5)
Anticonvulsants and antiparkinsonism drugs	966	17 026	2.4 (2.3-2.5)
Sedatives and hypnotics	967	26 805	3.8 (3.7-4.0)
Other central nervous system depressants and anesthetics	968	15 642	2.2 (1.9-2.6)
Psychotropic agents	969	169 309	24.2 (23.7-24.7)
Central nervous system stimulants	970	27 915	4.0 (3.5-4.6)
Drugs primarily affecting the autonomic nervous system	971	6710	1.0 (0.9-1.0)
Agents primarily affecting the cardiovascular system	972	20 878	3.0 (2.9-3.1)
Agents primarily affecting the gastrointestinal system	973	2616	0.4 (0.3-0.4)
Water mineral and Uric acid metabolism drugs	974	3133	0.5 (0.4-0.5)
Agents primarily acting on the smooth/skeletal muscles/respiratory	975	14 032	2.0 (1.9-2.1)
Agents primarily affecting skin/mucous membrane ophthalmologic	976	159	b0.1
otorhinolaryngologic and dental drugs
Other and unspecified drugs and medicinal substances	977	59 232	8.5 (8.0-9.0)
Bacterial vaccines	978	137	b0.1
Other vaccines and biological substances	979	145	b0.1
Other diagnosis codes not belong to poisoning	005-959, 980-999,	120 379	17.2 (16.4-18.1)
	V07-V68
Source: NEDS, 2007.

United States, the largest increases in fatal drug-related poisoning in the past 10 years have occurred in rural counties, and the current epidemic is driven by prescription drugs rather than illicit drugs. Our study’s contribution is the finding of a significantly higher rate of drug-related poisoning ED visits (most of these cases are nonfatal drug- related poisonings) in nonmetropolitan/micropolitan areas when compared with other areas. However, our study and the study by Paulozzi et al [6] could not provide direct evidence

Table 4 Percentage of unintentional and suicidal poisoning (%) by 2 major classes of drugs listed as first diagnosis, 2007

ICD-9 code National estimate

Weighted %

Unintentional

Suicidal Other ^a

Table 3 Drugs listed as first diagnosis responsible for US ED visits of drug-related poisoning, 2007

about underlying factors causing the significantly higher rates of fatal and nonfatal drug-related poisoning in rural areas than in urban areas. Studies have indicated that, in recent years, the use of prescription opioids and drug-related poisoning are more common in Rural communities than in urban communities in the United States [23,24]. Two factors have been proposed by researchers as possible causes for this geographic difference: less pain management experience of physicians in rural areas and urban-rural population

Analgesics, antipyretics, and antirheumatics Opiates and related narcotics Opium (alkaloids) unspecified Heroin Methadone Other opiates and related narcotics Salicylates Aromatic analgesics not elsewhere classified Psychotropic agents Antidepressants Benzodiazepine-based tranquilizers Other tranquilizers	965 965.0 965.00 965.01 965.02 965.09 965.1 965.4 969 969.0 969.4 969.1, 969.2, 969.3, 969.5	161 510 66 772 12 167 12 392 9380 32 832 10 311 57 758 169 309 40 510 79 260 28 931	40.1 49.8 49.6 45.2 53.7 50.4 32.2 32.2 29.5 25.3 27.5 27.7	41.1 22.8 18.2 11.8 18.8 29.8 55.5 56.1 52.4 61.9 54.1 59.4	18.8 27.4 32.3 43 27.5 19.9 12.3 11.7 18.1 13.3 18.9 13.3
Source: NEDS, 2007. a Among the “Other” category of intent, approximately 0.5% were assault cases, and approximately 99.5% were cases with unknown intent.

differences [6]. Physicians in rural areas may have less experience and training in using prescription opioids in their practice [25]. Pletcher et al [21] examined pain-related ED visits using data from the 1993-2005 National Hospital Ambulatory Medical Survey. Their study found that opioid prescribing rates for managing pain at US ED visits have increased substantially in recent years; however, nonwhite patients are less likely to receive an opioid analgesic than non-Hispanic white patients. Other researchers [26] also have suggested that pharmacies in some minority commu- nities are less likely to stock opioid analgesics, which may explain the lower use of these drugs in some urban communities. Although the exact underlying factors for a major shift in US drug-related poisonings in the past decade are not clear, the results of our study and other researchers suggest that prevention and control of drug-related poisoning warrant further research and that prevention strategies may need to shift focus to prescription opioid analgesics and populations not considered at high risk in the past.

Important strengths of our study include the large sample size of the NEDS and the examination of drug-related poisoning rate patterns by age, sex, urban-rural areas, and household income. In addition, the nature of the NEDS’ sampling scheme allows our results to be generalized to all drug-related poisoning ED visits in the United States However, our study has the following limitations. First, only the first-listed ICD-9-CM diagnosis code was used to identify drugs involved in these drug-related poisoning ED visits. Our analysis of the 2007 NEDS data (results not shown in the tables) indicated that cases with the first-listed ICD-9-CM code for a medical condition other than drug- related poisoning were significantly more likely to be admitted into the hospital (46.7%; 95% CI, 44.1%-49.4%) compared with cases with the first-listed ICD-9-CM code for drug-related poisoning (40.1%; 95% CI, 38.9%-41.3%), and the former cases were more likely to die in the ED (0.20% [95% CI, 0.15%-0.27%] vs 0.06% [95% CI, 0.04%-

0.08%]). These results suggest that mild or moderate poisoning cases were likely to have poisoning as their primary diagnosis, but Severe poisoning cases were likely to have a critical illness as the primary diagnosis. The full implication of this finding is unclear. Multiple-drug involvement is common in both intentional and uninten- tional drug-related poisoning. For visits with drug-related poisoning as the primary diagnosis, it is also likely that secondary diagnoses involved other drugs or substances that could have contributed to the poisoning. Thus, this study underestimated the potential contribution of drugs listed as secondary diagnoses. Second, it is likely that this study underestimated the number of drug-related poison- ings, because it only included cases treated in community, nonrehabilitation hospital-based EDs. The findings of this study may not be representative of drug-related poisoning cases treated in other health care settings or cases that did not receive treatment at all. Third, the NEDS does not capture all drug-related fatalities, because not all drug-

related poisoning deaths are transported to an ED. Finally, only the charges for treating Drug-related ED visits are available in the NEDS database. Unlike other Health Care Utilization Project data sets, there is no cost-to-charge ratio file available to convert charges into costs in this study. Therefore, true costs of these drug-related poisoning ED visits are unknown.

Conclusions

This study identified almost 700 000 visits to the United States hospital EDs due to drug-related poisoning in 2007. This underscores the importance of drug-related poisoning as a major public health problem in the United States. Our study adds new evidence that the current epidemic of drug-related poisoning is fueled by prescription opioids, and rural areas have higher rates of drug-related poisoning than urban areas. prevention programs should target prescription opioids, and rural areas, as well as children 0 to 5 years old, because they experience the highest rate of unintentional poisoning. Suicide-related poisoning should also receive attention, especially among females.

Acknowledgments

We thank Ms Krista Wheeler (Center for Injury Research and Policy, The Research Institute at Nationwide Children’s Hospital) for her great effort in editing the text during the revision of the manuscript.

References

1. Manchikanti L. National drug control policy and prescription drug abuse: facts and fallacies. Pain Physician 2007;10:399-424.
2. Dunn KM, Saunders KW, Rutter CM, et al. Opioid prescriptions for chronic pain and overdose: a cohort study. Ann Intern Med 2010;152: 85-92.
3. Gilson AM, Ryan KM, Joranson DE, et al. A reassessment of trends in the medical use and abuse of opioid analgesics and implications for diversion control: 1997-2002. J Pain Symptom Manage 2004;28: 176-88.
4. Centers for Disease Control and Prevention. Unintentional poisoning in the United States; 2010. p. 4.
5. Centers for Disease Control and Prevention. QuickStats: Death rates for the three leading causes of injury death-United States, 1979-2007. MMWR 2010;59:957.
6. Paulozzi LJ, Xi Y. Recent changes in drug poisoning mortality in the United States by urban-rural status and by drug type. Pharmacoepi- demiol Drug Saf 2008;17:997-1005.
7. Kuehn BM. Alarming nonfatal overdose rates found for opioids, sedatives, and tranquilizers. JAMA 2010;303:2020-1.
8. Berlow S, Paulozzi L, Diekman S. Prescription drug overdose: state health agencies respond; 2008.
9. Drug Abuse Warning Network (DAWN). National estimates of drug- related emergency department Visits; 2010. p. 104.
10. Centers for Disease Control and Prevention. Emergency department visits involving nonmedical use of selected prescription drugs- United States, 2004-2008. MMWR 2010;59:705-9.
11. Introduction to the HCUP Nationwide Emergency Department Sample 2007, Version 2. Agency for Healthcare Research and quality healthcare Cost and Utilization Project. Available http:// www.hcup-us.ahrq.gov/db/nation/neds/NEDS_Introduction_2007v2. jsp. 2010.
12. SAS Institute Inc: Statistical analysis software (SAS). Cary, NC: SAS Institute; 2004.
13. SUDAAN Inc. Software for the statistical analysis of correlated data. Research Triangle Park, NC: Research Triangle Institute; 2004.
14. NC-EST2007-ALLDATA-R-File16. US Census Bureau Monthly Postcensal Resident Population, by single year of age, sex race, and Hispanic origin: July 1, 2007. (http://www.census.gov/popest/ national/asrh/files/NC-EST2007-ALLDATA-R-File16.csv) (Accessed Oct 1, 2010).
15. Barrett M, Hunter K, Coffey R, et al. Population denominator data for use with the HCUP databases (Updated with 2009 Population Data). HCUP Methods Series Report # 2010-02. Online April 12, 2010. US Agency for Healthcare Research and Quality; 2010. Available: http:// www.hcup-us.ahrq.gov/reports/methods.jsp.
16. National Center for Injury Control and Prevention. WISQARSTM (Web-based Injury Statistics Query and Reporting System): Nonfatal Injury Data; 2010.
17. Bronstein AC, Spyker DA, Cantilena Jr LR, et al. 2007 Annual report of the American Association of Poison Control Centers’ National

Poison Data System (NPDS): 25th annual report. Clin Toxicol (Phila) 2008;46:927-1057.

1. Franklin RL, Rodgers GB. Unintentional child poisonings treated in United States hospital emergency departments: national estimates of incident cases, population-based poisoning rates, and product involvement. Pediatrics 2008;122:1244-51.
2. Schillie SF, Shehab N, Thomas KE, et al. Medication overdoses leading to emergency department visits among children. Am J Prev Med 2009;37:181-7.
3. Madden MA. Pediatric poisonings: recognition, assessment, and management. Crit Care Nurs Clin North Am 2005;17:395-404.
4. Pletcher MJ, Kertesz SG, Kohn MA, et al. Trends in opioid prescribing by race/ethnicity for patients seeking care in US emergency departments. JAMA 2008;299:70-8.
5. Kuehn BM. Medication helps make therapy work for teens addicted to prescription opioids. JAMA 2010;303:2343-5.
6. Paulozzi LJ, Ryan GW. Opioid analgesics and rates of fatal drug poisoning in the United States. Am J Prev Med 2006;31:506-11.
7. Centers for Disease Control and Prevention. Unintentional poisoning deaths-United States, 1999-2004. MMWR Morb Mortal Wkly Rep 2007;56:93-6.
8. Breuer B, Pappagallo M, Tai JY, et al. US board-certified pain physician practices: uniformity and census data of their locations. J Pain 2007;8:244-50.
9. Green CR, Ndao-Brumblay SK, West B, et al. Differences in prescription opioid analgesic availability: comparing minority and white pharmacies across Michigan. J Pain 2005;6:689-99.