SEARCH

SEARCH BY CITATION

Keywords:

  • prescription drug abuse;
  • poison center calls;
  • emergency department visits;
  • pharmacoepidemiology

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

Background

Prescription drug abuse is a critical problem in the USA and has been linked to more deaths than automobile accidents. Despite this growing epidemic, the USA lacks a timely early warning system. Poison centers (PCs) have the potential to act as sentinel reporting entities for prescription drug abuse and misuse due to near-real-time data reporting and abundant coverage in the USA.

Methods

Data from the Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS®) System PC program were compared with data from the Drug Abuse Warning Network (DAWN) from 2004 through 2010. Population rates of PC call mentions regarding abuse and misuse of prescription opioids were compared with population rates of emergency department visit mentions of the same using linear regression. Products included in the analysis were the following: buprenorphine, fentanyl, hydrocodone, hydromorphone, methadone, morphine, and oxycodone.

Results

The strength of association between RADARS System PC data and DAWN emergency department visits regarding all opioids in aggregate was strong (R2 = 0.81, p < 0.001). The correlations between the two programs at the drug class level also were strong for buprenorphine, hydrocodone, hydromorphone, methadone, and oxycodone (all R2 > 0.70, all p < 0.01), significant for fentanyl (p = 0.05), and moderate for morphine (p = 0.09).

Conclusions

Data on prescription opioid drug abuse from the RADARS System PC program correlates well with emergency room data from DAWN. Due to timeliness of data, geographic coverage and strong associations with other warning systems, PC data can be used for sentinel reporting on prescription drug abuse and misuse in the USA. Copyright © 2013 John Wiley & Sons, Ltd.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

Prescription opioid abuse has been increasing in recent years and is a topic of major concern throughout the USA. Emergency department (ED) visits resulting from prescription drug overdose have more than doubled from 2004 to 2010, while ED visits mentioning illicit drugs have remained relatively stable over the same period.[1] Deaths due to prescription opioid abuse and misuse increased threefold from 1999 to 2006,[2] and $8.6bn have been attributed to opioid abuse because of related medical expenses, criminal justice, and costs due to lost work.[3]

Public health interventions utilizing surveillance systems have the potential to reduce the national burden of prescription drug abuse and misuse. However, publication of data from large, federally-funded monitoring systems frequently is delayed a year or more. Unfortunately, it has been reported that decreases in funding, coupled with inefficiencies, have triggered the reorganization of the Drug Abuse Warning Network (DAWN). DAWN was dramatically downsized and will utilize only 730 ED visits instead of 350 000–400 000 it has historically used to generate nationwide estimates.[4] DAWN issued a final formal report pertaining to 2011 visits and a reduced form of DAWN will be operated within the National Center for Health Statistics through the Centers for Disease Control and Prevention.[4] A timely, geographically specific early warning system for the detection of prescription opioid abuse and misuse poisoning would likely increase the effectiveness of public health strategies by allowing specific targeted interventions as well as the evaluation of interventions. With the downsizing of DAWN, a vacuum in this regard has been created. Poison centers (PCs) can fill this role as they are capable of near-real-time electronic reporting of exposures to substances, including product-specific prescription opioid medications.[5] PC data have been used in post-marketing surveillance and evaluating policy changes,[6-9] and have been proven successful in predicting methadone-related mortality.[10] In addition to timeliness, PCs have considerable national coverage. Due to these strengths, PCs are a critical component of the non-profit Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS®) System.[11-13] Currently, PCs participating in the RADARS System PC program have over 90% nationwide coverage and collect data on the nature of the exposure, including the specific product and geographic location of the exposure.

The aim of this study was to further examine the ability of PCs to act as sentinel indicators of prescription drug abuse and misuse. Associations of population rates of drug abuse and misuse-related PC call mentions with population rates of ED visit mentions related to abuse and misuse documented by DAWN are presented. We hypothesized that emergency call data from PCs would correlate well with ED visit data from DAWN.

METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

Researched Abuse, Diversion and Addiction-Related Surveillance System Poison Center data

The RADARS System is a non-profit prescription drug abuse and misuse surveillance program and a part of the Rocky Mountain Poison and Drug Center, a division of the Denver Health and Hospital Authority. PCs receive spontaneous calls from caregivers, patients, and healthcare providers regarding potentially toxic exposures. PC specialists trained in toxicology assist in the care of the individual and document numerous aspects of each case, including medical outcome and the place where individuals with an exposure-related call are managed. Medical outcomes are defined by the American Association of Poison Control Centers coding manual as follows: (i) Minor Effect is an instance where the patient had minimally bothersome symptoms, such as drowsiness, where after a period of follow-up, the patient is determined to have returned to a pre-exposure state; (ii) Moderate Effect is progressively more severe, but not life-threatening, and recorded when the patient needed some type of medical treatment; and (iii) Major Effect is recorded when the patient had life-threatening symptoms or significant residual disability resulting from the exposure. PC specialists also document where the case was treated: Managed On-Site (not a healthcare facility), In Route to Healthcare Facility, or Referred to Healthcare Facility.[14]

Participating PC records are uploaded to a central database in the RADARS System where case review and quality control are conducted, and exposure mentions are recorded for analysis. Quality control includes a review to verify exposure substances, exposure reasons, and removal of identifying information.[15] Exposure call data also is linked to the three-digit zip code of the call location. The RADARS System has expanded its participating PC coverage from 40% of the USA population in 2004 to 85% of the population in 2010.

The Drug Abuse Warning Network

Drug Abuse Warning Network utilizes a network of EDs, which treat individuals and document visits mentioning drug exposures. DAWN provides information regarding a wide range of drug-related visits, including ones in which misuse or abuse of prescription opioids are noted. In 2009, 242 hospitals in 12 major metropolitan regions throughout the country contributed data to DAWN estimates. DAWN uses a complex survey design to generalize population rates of ED visit mentions to the entire population of the country.[16] More specifically, case mentions are generalized to the country through estimates that have weighted values based on participating EDs. While no ED reports exist from certain regions of the country, an estimated frequency of visits they would have reported is aggregated in the nationwide estimate, which is based on participating EDs and a complex statistical algorithm.

Populations included

Case mentions from 2004 through 2010 regarding the abuse and misuse mentions of prescription opioid medications resulting in various known medical outcomes were selected from the RADARS System PC program and DAWN. A mention refers to a drug being named by an individual while receiving care for the exposure; an individual can name numerous drugs during an emergent episode. Drug exposures included in this analysis were the following prescription opioids products (class level): buprenorphine, fentanyl, hydrocodone, hydromorphone, methadone, morphine, and oxycodone.

Cases mentions recorded in DAWN originate specifically from ED visits, compared with PC case mentions which originate from different locations and may involve a more broad range of medical effects. These variables may result in divergent abuse and misuse rate relationships among the programs. To evaluate these potential differences, cases from PCs were classified in different groups based on origin of call, and followed medical outcomes. Those missing medical outcomes were not included in the medical outcome analysis. The distribution of age groups in both programs is presented in Figure 2. This distribution is classified to match age categories presented by DAWN with 3.5% of cases in the RADARS System PC program lacking classifiable age information. Over the study period from 2004 through 2010, 160 614 PC program case mentions and more than 2.2 million estimated ED visits from DAWN were included in the analysis.

Statistical analysis

Population rates of case mentions regarding abuse and misuse of prescription opioids collected by the RADARS System PC program were compared with population rates of abuse and misuse case mentions from ED visits recorded by DAWN. From 2004 to 2012, participation in the RADARS System PC program more than doubled. Utilizing population rate data accounted for the growth of the RADARS System during the study period and allowed a direct comparison of DAWN rates. Population rates for PC program case mentions were calculated using the 2000 US Census population of the region covered by participating PCs as the denominator. Covered population was corrected linearly assuming a consistent 0.97% growth per year. Nationally, the population grew by 9.7% between the 2000 and 2010 census. Population rates were calculated annually for PC program data to match data available from DAWN. Population rates of case mentions for DAWN were obtained from the data file, which includes national estimates of all drug-related misuse and abuse ED visits through 2010 posted at http://www.samhsa.gov/data/DAWN.aspx. In both programs, population rates of prescription opioid mentions per 100 000 individuals were compared.

Associations between the two programs were evaluated based on PC program case calls grouped by management location (on-site versus healthcare facility). Second, PC program cases were classified in three different groups based on medical outcomes, and these classifications were used to generate different population rates to compare with DAWN. The groups included the following: (i) all PC exposure case calls with a followed medical outcome; (ii) only case calls where the individual had a no effect or minor effect medical outcome; and (iii) individuals who had more serious medical outcomes including; moderate, major, and death.

Linear regression with a t-test of significance was used to test associations of annual population rates between both programs. Although non-linear relationships have been identified between PC calls and methadone-related deaths,[10] a parsimonious linear relationship was anticipated due to a portion of PC calls originating from EDs. Given this is an analysis over consecutive time points; it may be subject to time-correlated observations. Potentially, the independence assumption of simple linear regression could be violated due to correlated observations in time. To account for this an error structure, correcting for time-correlated observations, by assuming a stronger correlation of points closer in time, was used when applicable. This was evaluated by comparing the Akaike information criterion score between regression models with and without a time correlation correction. R2 values unadjusted for autocorrelation and related p-values, adjusted for autocorrelation when appropriate, are presented.

First, a full regression model was developed to test associations over all drug classes combined. This included the DAWN population rate as the dependent variable and PC program population rate, drug, medical outcome group, and all two-level and three-level interaction terms as independent variables. The inclusion of interaction terms tested different slopes between the different programs by medical outcome groups. This was done to evaluate if the association between DAWN and the RADARS System PC program was dependent on medical outcome group classification of PC program data. Stratified analyses are presented in the case of a significant medical outcome group by rate interaction term.

Then, linear regression was used to evaluate associations of population rates between programs within class-specific prescription opioids. This included all PC program medical outcome groups, but the analysis was separated at the drug class level. This was done to evaluate the ability of PCs, regardless of medical outcome, to predict ED visits mentioning class level prescription opioids. Spearman correlation coefficients are presented as well to assess the correlation of the rank order of population rates of specific opioid drug classes. Analyses were conducted with sas Enterprise Guide v 4.3. (SAS Institute Inc., Cary NC, USA). The Poison Center Program protocol was granted approval by the Colorado Multiple Institutional Review Board (COMIRB). All participating poison centers obtain Institutional Review Board (IRB) approval and use trained health care professionals to collect call information and generate case notes on a standardized electronic data collection form.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

Table 1 lists the 11 major metropolitan regions of EDs participating in the DAWN program in 2010. In 2009, 242 hospitals within these regions (and additionally Houston, TX) contributed to DAWN population rate estimates. Complex survey methodology is used by DAWN to generalize rates nationwide. This contrasts with RADARS System data, in which the PC program captures case mentions from the majority of the country (Figure 1).

Table 1. Drug Abuse Warning Network participating major metropolitan regions 2010
Boston, MA.Minneapolis, MN.
Chicago, IL.New York City—5 Boroughs, NY.
Denver, CO.Phoenix, AS.
Detroit, MI.San Francisco, CA.
Miami—Dade, FL.Seattle, WA.
Miami—Fort Lauderdale, FL. 
image

Figure 1. The geographic coverage of Researched Abuse, Diversion and Addiction-Related Surveillance System poison center program in 2010, which accounted for 85% of the population

Download figure to PowerPoint

Medical outcomes of PC program cases are presented in Table 2. Notably, a substantial number of PC program calls have a minimal level of medical severity (56% no effect and minor effect) and are managed frequently from on-site not a health care facility (10%, data not shown). There were no significant differences of the association between programs based on the origin of call, therefore this analysis is not presented.

Table 2. Medical outcome characteristics of the poison center prescription opioid exposure case population examined
Medical outcomeN%
No effect20 38617.7%
Minor effect44 35238.4%
Moderate effect38 05332.9%
Major effect11 63510.1%
Death10680.92%
Total160 614100%

Women are slightly more frequent in PC program data relative to DAWN data: 55% of case mentions versus 47% of case mentions, respectively. In PC program data, the age distribution is shifted left slightly, indicating a younger population than that reported by DAWN (Figure 2).

image

Figure 2. The age distribution of individuals reported by the two programs

Download figure to PowerPoint

Due to the significant interaction of rate by group ( p < 0.001), stratified analyses of combined prescription opioids are presented by medical outcome group for ease of interpretation and graphically in Figure 3. Interestingly, opioids in aggregate have a strong association with PC calls in the minimal medical effect outcome group (F = 70, on 1 and 46 d.f., p < 0.001, R2 = 0.60); however, a clear difference is observed with hydrocodone (Figure 3, Panel A). In the group defined as minimal medical effect, hydrocodone is disproportionately overrepresented in PC program data relative to DAWN data. Despite this shift, population rates within hydrocodone are highly associated between the two programs (F = 74 on1 and 5 d.f., p < 0.001, R2 =0.94). The association of all opioids in aggregate between DAWN and the PC program defined by a more serious medical outcome is strong as well (F = 201, on 1 and 46 d.f., p < 0.001, R2 = 0.81); but in this case, a shift in hydrocodone is not evident (Figure 3, Panel B). In this group, for each 0.1/100 000 increase in population rate detected by the PC program, an average 15/100 000 increase in population rate was detected by DAWN (b = 151, 95% CI = 130–172, p < 0.001).

image

Figure 3. The association of population rates between programs stratified by poison center program medical outcome group. Panel A includes only no effect or minor effect cases, and panel B includes cases with moderate effect, major effect, or death. Rates per 100 000 population from Drug Abuse Warning Network are presented on the y-axis, and rates per 100 000 population from Researched Abuse, Diversion and Addiction-Related Surveillance System poison center program are presented on the x-axis

Download figure to PowerPoint

The Spearman correlation of the rank order of population rates for drug classes for both programs over the study period is presented in Table 3. The correlations per year are strong, ranging from 0.89 through 0.93. Only hydrocodone and oxycodone show regular differences between the two programs. Hydrocodone consistently is the highest in RADARS System PC program data, whereas oxycodone consistently is the highest in DAWN data.

Table 3. Spearman correlation values of ranked population rates (including all poison center cases) between both programs over time
YearSpearman's r
20040.93
20050.89
20060.89
20070.89
20080.93
20090.93
20100.89

Figure 4 displays linear associations segregated by specific drug classes. This analysis included all PC program abuse and misuse exposure case mentions with a recorded medical outcome. At the drug class level, associations were strong for most opioids and moderate for morphine and fentanyl. Associations ranged from (F = 4.4, on 1 and 5 d.f., p = 0.09, R2 = 0.46, morphine) to (F = 437, on 1 and 4 d.f., p < 0.001, R2 = 0.99, buprenorphine). The proportion of the variance in DAWN rate estimates explained by the RADARS PC program rate estimates regarding morphine and fentanyl is substantial (46% and 57%, respectively).

image

Figure 4. Class-specific linear associations between programs including all Researched Abuse, Diversion and Addiction-Related Surveillance System poison center program exposure calls with recorded medical outcomes. Rates per 100 000 population from Drug Abuse Warning Network are presented on the y-axis, and rates per 100 000 population from Researched Abuse, Diversion and Addiction-Related Surveillance are presented on the x-axis

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

Overall, strong associations of population rate mentions between the two programs were found, regardless of PC program case classifications. This finding suggests RADARS System PCs could be used as sentinel reporting entities and as an effective tool for predicting ED visits resulting from abuse and misuse of prescription opioids. The findings from this study complement earlier works demonstrating the utility of using PC data as a surveillance tool for monitoring the effects of non-medical use of prescription opioid medication,[17-19] nonprescription medications,[20-23] and methadone-related mortality.[10]

This study also highlighted the difference between the two programs with regard to the medical outcome classification of hydrocodone. Within hydrocodone, associations are strong regardless of medical outcome; however, an interesting displacement of hydrocodone relative to other opioids was observed when less severe medical outcomes from PCs were used for comparison. The data suggest hydrocodone, at a population level, may be a less dangerous opioid of abuse and misuse. However, this finding may imply the existence of a unique, sentinel population captured by PC program data. It is possible that this population has a higher percentage of novel opioid experimenters, therefore represents a group that may benefit from interventions. Additional research would be needed to study this question. Other studies also have identified different population characteristics in PC data versus other collection systems[10, 17, 24]; for example, a greater percentage of PC cases involve female patients and a younger population in general. Further, hydrocodone had the highest population rate detected in the PC program, while oxycodone had the highest rate detected in DAWN. Interestingly, Butler et al. reported that hydrocodone had the lowest risk of abuse in an opioid treatment population relative to other opioids, whereas oxycodone had an increased risk of abuse.[25] These findings may suggest that seasoned opioid abusers prefer oxycodone and other opioid medications over hydrocodone; however, the populations prescribed with hydrocodone are markedly different from those prescribed with other opioids. For example, patients prescribed with fentanyl may have additional comorbidities from conditions leading to chronic pain, and patients prescribed with buprenorphine are likely receiving it for addiction treatment. More research is needed to understand these population characteristics in order to verify a preference for hydrocodone by introductory and early misusers and abusers.

Prescription opioid class-specific level associations were strong for most opioids and moderate for morphine and fentanyl. While not statistically significant, the proportion of DAWN variance explained by the RADARS System PC program data regarding morphine was considerable, and additional time points may make that association significant. Further investigation regarding characteristics of morphine and fentanyl abusers could inform on these differences.

Approximately 18% of exposure-related calls received by PCs originate from healthcare facilities. This offers a partial explanation for the strong associations between programs; however, this also highlights the importance of PCs regarding prescription drug surveillance. PCs are able to detect abuse and misuse-related exposures that are missed by EDs. On the other hand, the strong associations identified between both programs could be indicative of the level of availability of prescription opioids in the USA. Kuehn noted increased prescriptions for opioid medication coincided with increased abuse.[26] Abuse and misuse of prescription drugs, whether resulting in an ED visit or PC call, could be suggestive of the availability of the drug chosen for abuse and misuse. With increasing availability, both ED visits and PC calls would be expected to increase, in accord with the findings presented by Kuehn.

There are limitations to this study. Only seven time points were available for comparison; therefore, this study would be strengthened by additional points. Further, despite raising questions regarding the characteristics of specific drug abusing and misusing populations, this study is unable to make causal inferences due to its ecological nature. Also, it should be noted that numerous calls to PCs are missing medical outcome information and the overall frequency abuse and misuse detected by DAWN is substantially larger. However, this study had adequate numbers for comparison with regard to medical outcome, allowing for stratified analyses resulting in strong associations. There are three offsetting strengths regarding the use of PCs as sentinel surveillance tools: (i) the persistent associations across groups, regardless of medical severity; (ii) the persistent associations across class-specific prescription opioid groups; and (iii) the availability of timely information.

CONCLUSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

Drug Abuse Warning Network is a respected and well-known drug abuse surveillance system. The RADARS System PC program data agree well with DAWN, regarding drug abuse and misuse trends. The RADARS System is designed to capture accurate near-real-time abuse and misuse data with product and geographic specificity. The RADARS System PC program has coverage across the majority of the US population and collects geographic information directly without complex statistical inference.

Importantly, beyond correlating with DAWN data, this study demonstrated that PCs can capture potentially different populations than other medical surveillance systems. Furthermore, PCs appear to have heightened sensitivity and quick response time, which may help to detect emerging drug abuse and misuse problems. PCs, therefore, are a critical national component in drug abuse and misuse surveillance and can provide timely, geographically specific, and highly actionable information.

CONFLICT OF INTEREST

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

The authors are current or former employees of Denver Health and Hospital Authority, which operates the Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS®) System. Several manufacturers of controlled substances are subscribers to the RADARS System. No authors have direct financial relationship with any of these companies.

KEY POINTS

  • PC calls have a strong association with ED visits.
  • PCs can act as sentinel drug abuse and misuse reporting entities.
  • PCs also detect abuse in populations with minimal medical side effects.

ETHICS STATEMENT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES

This research was reviewed and approved by the Colorado Multiple Institutional Review Board and the Denver Health and Hospital Authority, in addition to the local institutional review boards for all participating poison centers. Data were obtained in de-identified format analyzed anonymously.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ETHICS STATEMENT
  10. REFERENCES
  • 1
    Substance Abuse and Mental Health Services Administration, Center for Behavioral Health Statistics and Quality. The DAWN Report:Highlights of the 2010 Drug Abuse Warning Network (DAWN) Findings on Drug-Related Emergency Department Visits. Rockville, MD. 2012.
  • 2
    Warner M, Chen LH, Makuc DM. Increase in Fatal Poisonings Involving Opioid Analgesics in the United States, 1999–2006. NCHS data brief, no 22. Hayattsville, MD: National Center for Health Statistics. 2009.
  • 3
    Birnbaum HG, White AG, Reynolds JL, et al. Estimated costs of prescription opioid analgesic abuse in the United States in 2001: a societal perspective. Clin J Pain 2006; 22(8): 667676.
  • 4
    Knopf A (editor). DAWN report for 2011 will be the last; transition to CDC. Alcohol Drug Abuse Wkly 2012; 24(38): 35.
  • 5
    Hughes AA, Bogdan GM, Dart RC. Active surveillance of abused and misused prescription opioids using poison center data: a pilot study and descriptive comparison*. Clin Toxicol 2007; 45(2): 144151.
  • 6
    Bunn TL, Yu L, Spiller HA, Singleton M. Surveillance of methadone-related poisonings in Kentucky using multiple data sources. Pharmacoepidemiol Drug Saf 2010; 19(2): 124131.
  • 7
    Dart RC, Dasgupta N, Bailey JE, Spiller HA. Interpreting poison center call volume associated with tramadol. Ann Pharmacother 2011; 45(3): 424.
  • 8
    Tormoehlen LM, Mowry JB, Bodle JD, Rusyniak DE. Increased adolescent opioid use and complications reported to a poison control center following the 2000 JCAHO pain initiative. Clin Toxicol 2011; 49(6): 492498.
  • 9
    Spiller HA, Scaglione JM, Aleguas A, et al. Effect of scheduling tramadol as a controlled substance on poison center exposures to tramadol. Ann Pharmacother 2010; 44(6): 10161021.
  • 10
    Dasgupta N, Davis J, Jonsson Funk M, Dart R. Using poison center exposure calls to predict methadone poisoning deaths. PLoS ONE 2012; 7(7): e41181.
  • 11
    Schneider MF, Bailey JE, Cicero TJ, et al. Integrating nine prescription opioid analgesics and/or four signal detection systems to summarize statewide prescription drug abuse in the United States in 2007. Pharmacoepidemiol Drug Saf 2009; 18(9): 778790.
  • 12
    Cicero TJ, Dart RC, Inciardi JA, Woody GE, Schnoll S, Muñoz A. The development of a comprehensive risk-management program for prescription opioid analgesics: Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS®). Pain Med 2007; 8(2): 157170.
  • 13
    Available at: http://www.radars.org/. [1 August 2013]
  • 14
    American Association of Poison Control Centers. AAPCC National Poison Data System (NPDS) Reference Manual. American Association of Poison Control Centers: Washington, DC, 2007.
  • 15
    Smith MY, Dart R, Hughes A, et al. Clinician validation of poison control center (PCC) intentional exposure cases involving prescription opioids. Am J Drug Alcohol Abuse 2006; 32(3): 465478.
  • 16
    SAMHSA. Drug Abuse Warning Network, 2009: National Estimates of Drug-Related Emergency Department Visits. Substance Abuse and Mental Health Services Administration: Rockville, MD, 2011.
  • 17
    Vassilev ZP, Marcus S, Jennis T, Ruck B, Swenson R, Rego G. Rapid communication: sociodemographic differences between counties with high and low utilization of a regional poison control center. J Toxicol Environ Health A 2003; 66(20): 19051908.
  • 18
    Forrester MB. Oxycodone abuse in Texas, 1998–2004. J Toxicol Environ Health A 2007; 70(6): 534538.
  • 19
    Marquardt KA, Alsop JA, Albertson TE. Tramadol exposures reported to statewide poison control system. Ann Pharmacother 2005; 39(6): 10391044.
  • 20
    Bryner JK, Wang UK, Hui JW, Bedodo M, MacDougall C, Anderson IB. Dextromethorphan abuse in adolescence: an increasing trend: 1999–2004. Arch Pediatr Adolesc Med 2006; 160(12): 12171222.
  • 21
    Forrester MB. Methylphenidate abuse in Texas, 1998–2004. J Toxicol Environ Health A 2006; 69(12): 11451153.
  • 22
    Banerji S, Anderson I. Abuse of Coricidin HBP cough & cold tablets: episodes recorded by a poison center. Am J Health Syst Pharm 2001; 58(19): 18111814.
  • 23
    Spiller HA, Krenzelok EP. Epidemiology of inhalant abuse reported to two regional poison centers. Toxicol Clin Toxicol 1997; 35: 167173.
  • 24
    Sims SA, Snow LA, Porucznik CA. Surveillance of methadone-related adverse drug events using multiple public health data sources. J Biomed Inform 2007; 40(4): 382389.
  • 25
    Butler SF, Black RA, Cassidy TA, Dailey TM, Budman SH. Abuse risks and routes of administration of different prescription opioid compounds and formulations. Harm Reduct J 2011; 8(1): 29.
  • 26
    Kuehn BM. Opioid prescriptions soar: increase in legitimate use as well as abuse. JAMA 2007; 297(3): 249251.