Embed Size (px)
Citation preview
Clinical Toxicology, 36(1&2), 11-17 (1998)
Opioid Toxicity Recurrence After an Initial Response to Naloxone
William A. Watson; Mark T. Stele; Robert L. Muelleman; Micheal D. Rush
University of Missouri-Kansas City; Truman Medical Center, Kansas City, Missouri
ABSTRACT
Obiective: To determine the frequency and potential predictors of opioid toxi- city recurrence after a response to naloxone in adult Emergency Department patients. Methods: A retrospective case-control study of naloxone-treated patients with opioid toxicity over an 8-year period. Both the patient response to naloxone and recurrence of opioid toxicity was determined by an expert Delphi Panel. The frequency of opioid toxicity recurrence was compared by the duration of opioid effect, the route of opioid exposure, and the presence of other CNS depressant drugs. Results: Ninety of 221 (41%) cases with a discharge diagnosis of opioid toxicity were treated with naloxone; six patients were excluded because of a lack of toxicity. There was a response to naloxone in 50% of the 84 cases, and recurrence of toxicity in 31% (95% CI 17-496) of naloxone responders. The most common opioids were codeine, heroin, propoxyphene, and oxycodone/hydrocodone. Recurrence of toxicity was more common with long-acting opioids (p = 0.04), and was not associated with the route of opioid exposure (p = 0.42), or presence of ethanol and other CNS depressants (p 1 0.87). Conclusion: Opioid toxicity recurrence after a response to naloxone occurred in approximately 1/3 of adult Emergency Department opioid overdose cases. Recurrence was more common with long- acting opioids and was not associated with the route of opioid exposure. Other clinically useful predictors of toxicity recurrence were not identified.
Correspondence: Dr. William Watson, Department of Emergency Medicine, Truman Medical Center, 2301 Holmes Street, Kansas City, MO 64108. Tel: 816/556-3112; Fax: 816/881-6282; E-mail: [email protected]
11
Copyright (D 1998 by Marcel Dekker, Inc.
Clin
ical
Tox
icol
ogy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Nyu
Med
ical
Cen
ter
on 0
5/12
/13
For
pers
onal
use
onl
y.
12 Watson et al.
INTRODUCTION
Naloxone has a shorter duration of action than most op io id~. ’ -~ The short duration of action is an important consideration in Emergency Department (ED) treatment of opioid overdoses with naloxone. Recurrence of opioid toxicity after a response to naloxone could result in significant morbidity or death. While it is generally stated that recurrence of opioid toxicity after a response to naloxone is common, the frequency has not been described for most opioids.’t5-” In 98 cases of IV heroin toxicity treated with naloxone, recurrence of toxicity was not identified; however, 40% of the patients were observed for less than 2 hours.’ In one series of 10 cases of methadone toxicity and another series with 9 cases of toxicity secondary to other opioids, the recurrence of toxicity was described only as com- mon.218 This information, with only a few different opioids, does not provide a clinically useful estimate of the frequency of opioid toxicity recurrence in most EDs, where toxicity from many different opioids is treated. While heroin toxicity is common- ly treated in many urban EDs, toxicity secondary to codeine, propoxyphene, and oxycodonelh ydrocodone are also frequently reported. 13, l 4
The purpose of this study was to determine the frequency of opioid toxicity recurrence after a re- sponse to naloxone in sequential adult ED patients with toxicity caused by various opioids. Character- istics of the clinical findings, opioid involved, and naloxone therapy were evaluated as potential predictors of opioid toxicity recurrence.
METHODS
This retrospective case control study included all patient visits to one ED with a discharge diagnosis of opioid toxicity and treatment with naloxone. Cases were identified by searching our emergency department database for patients with a discharge diagnosis of toxicity secondary to heroin, or other opioid dependence (ICD-9-CM codes 304.0, 305.5), or opioid poisoning/overdose (965.0 to 965.09) between August 1 , 1987 and October 3 1 , 1995. The database searched includes all patient visits (n = 324,887) to the Truman Medical Center ED, an urban, university-affiliated hospital treating adult patients in Kansas City, Missouri.” Cases were
enrolled in the study when both opioid toxicity and naloxone administration were documented. The pro- tocol was approved by the University of Missouri- Kansas City Adult Health Sciences Institutional Review Board.
Each ED record was reviewed and data recorded on a pretested data collection form. Reviewers were trained to provide consistency in data collection, and definitions for the data collected were created. There were periodic meetings of the data collectors to review coding guidelines and resolve questions. l6 All data were reviewed for accuracy and consistency by one investigator Y A W ) and entered into dBase IV software. Data collected were patient demo- graphics, ED presentation, diagnostic evaluation, treatment, and disposition. Toxicity data included history of exposure to opioids and other drugs, and clinical evidence of toxicity. The naloxone dose(s) administered, route, documentation of response, and clinical course after administration were recorded.
A nine member expert Delphi Panel consisting of board-certified clinical toxicologists, medical toxicol- ogists, and emergency ph sicians evaluated the case information (Appendix). They were provided with documentation from the medical record about each patient’s mental status and respiratory status. The panel members were asked to determine if there was a response to naloxone, and whether the response was followed by the recurrence of opioid toxicity. Panel members were not provided with any infor- mation which would bias their preexisting definition of either event. It was expected that changes in mental status and respiratory rate or depth would be the primary factors that panel members would use to determine whether either outcome occurred.
Panel members categorized their evaluation of the response and recurrence for each case as either definite, probable, or indeterminate. A second iter- ation was undertaken of cases when the Delphi Panel was not in agreement. It was determined a priori that a definite outcome required agreement by at least eight panel members. For an outcome to be probable, no more than one panel member could define the case as indeterminate. Cases which were categorized as having either a definite or probable outcome by the Delphi Panel were defined as having either a response or opioid toxicity recurrence. Delphi Panel members were not provided with cases where either there was no documentation of the
Clin
ical
Tox
icol
ogy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Nyu
Med
ical
Cen
ter
on 0
5/12
/13
For
pers
onal
use
onl
y.
Opioid Toxicity Recurrence 13
patient’s status after naloxone administration or it was stated in the record that there was no response to naloxone.
The primary study outcome was recurrence of opioid toxicity after a response to naloxone. Those cases which responded to naloxone were stratified based on whether or not recurrence of opioid toxicity occurred. The groups were compared to determine whether there were any variables documented before naloxone administration which were associated with the recurrence of opioid toxicity. The primary vari- ables for comparison were the duration of action of the opioids involved, the presence of other CNS de- pressant drugs including ethanol, patient demograph- ics, and the reason and route of opioid exposure.
The frequency of response to naloxone and opioid toxicity recurrence are reported with 95 % confidence intervals. For statistical analysis the opioids were categorized as either long-acting (methadone, sustained-release morphine, propoxyphene) or short- acting (heroin, codeine, oxycodone, hydrocodone, hydromorphine, meperidine, morphine, pentazocine). Cases with more than one opioid were assigned to the category of the longest acting opioid present. A 2 x 2 Chi-square analysis was used to compare patient demographics and toxicity between the two groups. The Mann-Whitney U test was used to compare age and the naloxone dose administered before a response. Statistical significance was defined as alpha C 0.05.
RESULTS
During the study period a total of 221 cases with an ED discharge diagnosis of opioid toxicity were identified and reviewed. Ninety cases were treated with naloxone. Six patients were excluded because there was no evidence of opioid toxicity before naloxone administration; the 84 remaining patients made up the study sample. The study sample was 33 f 11 years of age, 42% female, 54% Caucasian, and 37% African-American. The reason for opioid exposure was documented to be suicide in 48%, drug abuse in 32%, chronic analgesic therapy in 1 1 %, and unknown in 9%. Four patients presented to the ED twice during the study period, with visits separated by at least 3 days.
There was documentation of coma in 49% of patients, and lesser decreases in mental status in the
remaining patients prior to naloxone administration. Pupil size was documented as 2 mm or less in 13% of patients. In 27% of cases, there was also docu- mentation of decreased respiratory rate or depth, and 21% of the cases were endotracheally intubated before naloxone administration. A history of opioid exposure was available in 90% of cases, and past opioid abuse was noted in 27% of cases.
Ten different opioids were identified (sustained- release morphine was categorized separately from other morphine formulations). A single opioid was involved in 82% of cases, and in 4 cases methadone was combined with a short-acting opioid. In 68 patients, the opioids were ingested orally, and IV use was documented in 13 cases. One case reported smoking, and 2 cases reported insufflation of heroin. In 7 patients, an opiate-positive urine drug screen was available, but there was no documentation of which drug was involved. The majority of cases (66%) received a single dose of naloxone, and 2.0 mg was the most common naloxone dose adminis- tered. Naloxone was administered as IV boluses, except for one dose given via an endotracheal tube and one dose intramuscularly. All cases who ingested opioids or other drugs orally were treated with ipecac syrup, activated charcoal, gastric lavage, or a combination of activated charcoal and gastric lavage.
Forty-two (50% ; 95 % CI 35-65 %) of the 84 cases had either a definite (n = 18) or probable (n = 24) response to naloxone, and 22 cases were categorized as indeterminate by the Delphi Panel. No response to naloxone was documented in 17 cases, and 9 cases did not have documentation. Documentation of the response usually included information about im- provement in the patient’s mental status and respi- ratory function. The response to naloxone was opioid withdrawal in 5 patients. There was recur- rence of opioid toxicity in 13 of 42 cases (31 %; 95% CI 17-45%) which was defined by the Delphi Panel as definite in 4 and probable in 9 cases. No recurrence of opioid toxicity was documented in 18 cases, and the available documentation was categorized as indeterminate by the Delphi Panel in the remaining 1 1 cases which responded to naloxone. Recurrence of opioid toxicity was usually described as lethargy or somnolence, and two cases had a decrease in respiratory rate or depth. Six of the patients who developed recurrence of opioid toxicity
Clin
ical
Tox
icol
ogy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Nyu
Med
ical
Cen
ter
on 0
5/12
/13
For
pers
onal
use
onl
y.
14 Watson et al.
were treated with additional naloxone, including two cases that had a continuous naloxone infusion started during their ED treatment.
Cases who had recurrence of opioid toxicity after a response to naloxone were similar in age, sex, the reason for opioid exposure, and extent of toxicity compared to those in whom there was no recurrence (Table 1). Patients most frequently received 2.0 mg of naloxone before a response was documented (range 0.2 mg to 4.0 mg), with similar doses administered before documentation of a response between the two groups (p = 0.60). Cases were in the ED for a similar duration of time (6.0 f 2.5 hours) regardless of whether there was recurrence of opioid toxicity. Patients who had recurrence of opioid toxicity were more often hospitalized (54%) compared to those who responded to naloxone and had no recurrence of toxicity (21 76, p = 0.08).
Cases of toxicity caused by long-acting opioids were more likely to have recurrence of opioid toxi- city than cases of toxicity secondary to short-acting opioids (p = 0.04, Table 2). No difference in the frequency of response (p = 0.65) or recurrence (p = 0.42) of opioid toxicity was seen when the IV and oral routes of opioid exposure were compared. Recurrence of opioid toxicity occurred with all opioids except hydromorphone, meperidine, and pentazocine. Recurrence of opioid toxicity was seen between 3 and 120 minutes after the initial response to naloxone.
DISCUSSION
These results provide an estimate of the frequency of recurrence of opioid toxicity in adult ED patients who present to the ED because of toxicity secondary to various opioids. The strengths of this data, compared to previous reports, are the number of patients, the different opioids involved, and the Delphi method of defining study outcomes. Past reports of opioid recurrence after a response to naloxone have either focused on one drug, such as methadone or heroin, or been single case reports.899918-21 The only previous series with different opioids included 9 patients with toxicity secondary to either dipipanone (n = 3), meperidine (n = 2), heroin (n = l), pentazocine (n = l), or dihydrocodeine (n = 2). There was no documen- tation in this series of how frequently recurrence of
opioid toxicity occurred after the response.2 Our series included all cases of opioid toxicity treated over an %year period, with toxicity secondary to 10 different opioids.
It is difficult to evaluate outcomes when there is no agreed-upon definition. In a prospective evalua- tion there would probably be general agreement on what constitutes complete reversal of profound CNS depression, respiratory depression, and constricted pupils by naloxone. There would likely be disagree- ment when the CNS depression was less profound or the response less than complete. This difficulty in defining clinical response is magnified when clinical observations from medical records are used. The Delphi method is well suited for determining outcomes that cannot be measured by quantitative means and removes the bias of a single individual.” By not providing the panel members information on how they should define the outcomes, we were un- able to state which of the factors was most important in their categorizing the cases. It would be expected that level of consciousness was most significant.
The duration of effect of the opioid producing toxicity appears to be the primary variable predicting whether recurrence of toxicity will develop after a response to naloxone. Recurrence of opioid toxicity was more common with methadone, sustained-release morphine, and dextropropyphene. Long-acting opi- oids made up 20% of the study cases and 54% of recurrent toxicity. Recurrence in the remaining cases, however, could not be predicted based on the duration of action of the drug involved (morphine, heroin, codeine, oxycodone, hydrocodone, meper- idine). The dose of naloxone required to produce a response was not clinically useful in predicting who would have recurrence of toxicity. This was expected, as there are no controlled data suggesting that the duration of response to naloxone is dose-
There was inadequate documen- tation of information about the dose’of opioid and time of exposure for our cases, and these factors could not be considered as potential predictors of toxicity recurrence.
Recurrence of opioid toxicity was primarily documented as CNS depression described as either lethargy or somnolence. There was no evidence, however, of a longer duration of time spent in the ED when toxicity recurred. Patient time in the ED would not be expected to be a good indicator of
Clin
ical
Tox
icol
ogy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Nyu
Med
ical
Cen
ter
on 0
5/12
/13
For
pers
onal
use
onl
y.
Opioid Toxicity Recurrence
Tablc I
15
Comparison of Patient Demographics, Concurrent Ingestions, and Toxicity Prwr to Naloxone Adminktratwn
Recurrence of No Recurrence of Opioid Toxicity Opioid Toxicity
(n = 13) (n = 29) P
Sex (MIF)
Route of Opioid Administration Age (Y)
Oral IV
Suicide Abuse
Unknown Respiratory
Depression Coma Concurrent Depressant
Ethanol Hospitalization
Reason for Exposure
Therapy
Ingestion
815 33 f9
11 (85%) 2 (15%)
3 (23%) 7 (54%) 1 (8%) 2 (15%)
4 (31%) 6 (46%)
9 (69%) 8 (62%) 7 (54%)
1811 1 35f16
23 (79%) 11 (38%)
12 (41 %) 12 (41 %) 4 (14%) 1 (3%)
8 (28%) 11 (38%)
19 (66%) 17 (59%) 6 (21 %)
0.76 0.78
0.42
0.49
0.87 0.87
0.90 0.87 0.08
morbidity since it is impacted by factors including the availability of hospital beds, toxicity of concurrently ingested drugs, and other variables which determine how long a patient is observed. Patients who had recurrence of toxicity, however, were more likely to be hospitalized.
The most important clinical complication of the recurrence of opioid toxicity is decreased respiratory function. Respirator arrest has been reported by various authors.s~'s* ' In one patient, recurrent respiratory arrest occurred approximately 47 hours after ingestion, approximately 3.5 hours after a continuous naloxone infusion was discontinued. * Two patients in this series with toxicity recurrence had documentation of decreased respiratory function that did not require specific therapy.
The study sample from our urban teaching hospital reflects the different types of opioid toxicity commonly treated in midwestern EDs. The most
commonly identified opioids in our series were codeine, heroin, propoxyphene, and oxycodonel hydrocodone. These are also the most common opioids reported to the Drug Abuse Warning Net- work (DAWN) and the Toxic Exposure Surveillance System (TESS). 1 3 7 1 4 There are significantly fewer cases of heroidmorphine toxicity treated in our ED than would be predicted by the DAWN data for a central city ED. The database allowed us to insure that all cases with a diagnosis of opioid toxicity upon discharge from the ED were included in the evalu- ation. In this series, naloxone was administered to 41 % (95% CI 3547%) of identified cases of opioid toxicity. The 1995 TESS data suggests that naloxone is administered to approximately 50% of patients. l4
The primary limitation of this study is the lack of an available definition of response to naloxone and the recurrence of opioid toxicity. In order to minimize investigator bias of the medical record
Clin
ical
Tox
icol
ogy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Nyu
Med
ical
Cen
ter
on 0
5/12
/13
For
pers
onal
use
onl
y.
16 Watson et al.
Table 2
Opioid Exposure and Recurrence Ajier a Response to Naloxone
Recurrence of No Recurrence of Opioid Toxicity Opioid Toxicity
(n = 13) (n = 29)
Long-acting Opioids 7 Methadone 4 Sustained-release 1
Morphine Propoxyphene 2
Short-acting Opioids 6 Heroin 2 Codeine 2 Oxycodonel
Hydrocodone 1 Hydromorphine 0 Meperidine 0 Morphine 0 Pentazocine 0 UDS Positive
for Opiates 1
5 2 1
2 24
8 7
3
On a 2 x 2 Chi-square analysis of long-acting vs short- acting opioids, p = 0.04.
data, an expert Delphi Panel was used to determine whether the primary study outcomes occurred. This method has been recommended for outcomes which have not been clearly defined. While a prospective experimental study design could utilize consistent definitions of a response to naloxone and recurrence of opioid toxicity, it is unlikely that such a trial will be undertaken to answer these questions. A number of methods were utilized to enhance the validity and reliability of the data collected.16 The relatively small sample size prevents an evaluation of the frequency of opioid toxicity recurrence for individual opioids. Opioids are a small portion (3 %) of our ED cases involving drugs of abuse, with cocaine (> 2000 cases) and phencyclidine (> 500 cases) being much more frequent during the study period. Opioids were approximately 6% of all overdoses and poisonings during the study period.
Medical record documentation was adequate for establishing the occurrence of the outcomes in this study. Ninety percent of the records had information about a response to naloxone and 74% had infor- mation to determine whether recurrence of opioid toxicity occurred. The duration of the study period is unlikely to have introduced a time bias into this study. We cannot identify specific changes in the ED or faculty practice use of naloxone or evaluation and treatment of drug toxicity which would impact the study outcome. As part of the core-content education of emergency residents, an annual discussion of antidote use in toxicology has been presented to emergency medicine residents by one faculty (WAW) since 1988. The discussion focuses on naloxone as a therapeutic agent rather than a diagnostic tool and did not change over the study period. The diagnostic use of naloxone in patients without a diagnosis of opioid toxicity was discouraged throughout the study period.
The results of this evaluation suggest that the frequency of opioid toxicity recurrence is approxi- mately 20% to 45% after an initial response to naloxone. While recurrence of toxicity is more frequent with long-acting opioids, it also occurs with short-acting opioids including heroin and codeine. There were no clinically useful predictors of which patients would have recurrence of toxicity after an initial response to naloxone.
ACKNOWLEWEMENTS
Melissa Balk, PharmD, Joanne Luong, PharmD, and Ruth Noland, PharmD, are acknowledged for reviewing the emergency records. This study was supported by a grant from Ohmeda Pharmaceutical Products Division, Inc., Liberty Corner, New Jersey.
REFERENCES
1. Weisman RS. Naloxone. In: Goldfank’s Toxicologic Emergencies, 5th ed. Goldfrank LR, Flomenbaum NE, Lewin NA, Weisman RS, Howland MA, Hoffman RS, eds., Norwalk, CT: Appleton & Lange,
Evans LE, Swainson CP, Roscoe P, Prescott LF. Treatment of drug overdosage with naloxone, a specific narcotic antagonist. Lancet 1973;1:452-455.
3. Berkowitz BA. The relationship of pharmacokinetics to pharmacologic activity: Morphine, methadone and
1994~784-786. 2.
Clin
ical
Tox
icol
ogy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Nyu
Med
ical
Cen
ter
on 0
5/12
/13
For
pers
onal
use
onl
y.
Opioid Toxicity Recurrence 17
4.
5 .
6.
7.
8.
9.
10.
11.
12.
13.
14.
IS.
naloxone. Clin Pharmucokinet 1976;1:219-230. Evans JM, Hogg MI, LUM JN, Rosen M. Degree and duration of reversal by naloxone of effects of morphine in conscious subjects. Br Med J 1974;2:
Ford M, Hoffman RS, Goldfrank LR. Opioids and designer drugs. Emerg Med Clin North Am 1990;8:
Chamberlain JM, Klein BL. A comprehensive review of naloxone for the emergency physician. Am J Emerg Med 1994;12:650-660. Hoffman RS, Goldfrank LR. The poisoned patient with altered consciousness. Controversies in the use of a ‘coma cocktail’. JAMA 1995;274:562-569. Buchner LH, Cimino JA, Raybin HW, Stewart B. Naloxone reversal of methadone poisoning. N Y State J Med 1972;72:2305-2309. Smith DA, h a k e L, Loflin JR, Yealy DM. Is admission after intravenous heroin overdose necessary? Ann Emerg Med 1992;21: 1326-1330. Osterwalder JJ. Naloxone for intoxications with intravenous heroin and heroin mixtures-harmless or hazardous? A prospective clinical study. J Toxicol Clin Toxicol 1996;34:409-416. Hoffman JR, Schriger DL, Luo JS. The empiric use of naloxone in patients with altered mental status: A reappraisal. Ann Emerg Med 199 1;20:246-252. Yealy DM, Paris PM, Kaplan RM, Heller MB, Marini SE. The safety of prehospital naloxone administration by paramedics. Ann Emerg Med
Data from the Drug Abuse Warning Network, Annual Emergency Room Data, 1992. Series I, Number 12- A. US Department of Health and Human Services publication no (SMA) 94-2080. Substance Abuse and Mental Health Services Administration. Litovitz TL, Felberg L, White S, KleinSchwartz W. 1995 Annual report of the America Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 1996;14:487-537. Watson WA, Stratton NA, Salomone JA Ill. Ecodes from emergency department records. Acad Emerg Med 1995;2:569-570.
589-591.
495-5 1 1.
1990;19:902-905.
16. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med
Medical assessment by a Delphi group opinion technique. N Engl J Med 1973;288:1272-1275.
18. Hendra TJ, Gerrish SP, Forrest AR. Fatal methadone overdose. BMJ 1996;313:481-482.
19. Gourlay GK, Coulthard K. The role of naloxone infusions in the treatment of overdoses of long half- life narcotic agonists: Application to normethadone. Br J Clin Pharmacol 1983;15:269-271.
20. Gober AE, Kearns GL, Yokel RA, Danziger L. Repeated naloxone administration for morphine overdose in a l-month-old infant. Pediatrics 1979;
21. Romac DR. Safety of prolonged, high-dose infusion of naloxone hydrochloride for severe methadone overdose. Clin Pharm 1986;5:251-254.
22. Longnecker DE, Grazis PA, Eggers GW Jr. Naloxone for antagonism of morphine-induced respiratory depression. Anesth Analg 1973;52: 447-453.
23. Glass PS, Jhaveri RM, Smith LR. Comparison of potency and duration of action of nalmefene and naloxone. Anesth Analg 1994;78:536-541.
1996 ;27: 305-308. 17. Milholland AV, Wheeler SG, Heieck JJ.
63 :606-608.
APPENDIX DELPHI PANEL MEMBERS
Jeffrey Brent, MD, PhD, Denver, Colorado Brent R. Ekins, PharmD, Fresno, California Judd E. Hollander, MD, Stony Brook, New York Debra Kent, PharmD, Vancouver, British Columbia S. Rutherfoord Rose, PharmD, Charlotte,
Jay Schauben, PharmD, Jacksonville, Florida Michael Shannon, MD, Boston, Massachusetts Knox H. Todd, MD, MPH, Atlanta, Georgia Richard S. Weisman, PharmD, Miami, Florida
North Carolina
Clin
ical
Tox
icol
ogy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Nyu
Med
ical
Cen
ter
on 0
5/12
/13
For
pers
onal
use
onl
y.
