Ultrasound Visualization of Ingested Tablets:A Pilot Study

Sean Patrick Nordt, M.D., Pharm.D., Colleen Campbell, M.D., Anthony Medak, M.D.,Christian Tomaszweski, M.D., M.S., and Richard F. Clark, M.D.

Study Objective. To evaluate the ability of ultrasonography as a radiographictechnique to visualize ingested tablets in the human stomach.

Design. Prospective, ultrasonographer-blinded, pilot study.Setting. Tertiary care medical center.Subjects. Fifteen healthy adults (12 male, 3 female; mean age 35.5 yrs).Intervention. After fasting for at least 12 hours, subjects were randomized to

either ingest a total of 10 tablets (four acetaminophen, four ibuprofen, andtwo multivitamin tablets) with 500 ml of water or to drink 500 ml of wateralone (control group).

Measurements and Main Results. Nine subjects were randomized to thegroup taking the tablets and six to the control group. Within 30 minutesafter ingesting the tablets with water or the water alone, each subjectunderwent ultrasonography 2 times by two respective blinded ultrasono-graphers (emergency department physicians who were experienced with theultrasound technique). Each performed the radiographic technique inisolation of the other ultrasonographer so as not to influence each other’sinterpretation. They recorded whether or not they thought tablets werevisible. One ultrasonographer was correct for 14 (93%) of the 15 subjects,but the other ultrasonographer was correct for only 6 subjects (40%).Interobserver agreement was assessed, with no agreement found betweenthe two ultrasonographers (Cohen � coefficient –0.25 [95% confidenceinterval –0.7–0.2]). The ultrasonographers also found six false positives (nodrugs present in the stomach but incorrectly identified as present by theultrasonographer) and four false negatives (tablets present in the stomachbut not identified by the ultrasonographer).

Conclusion. Our results do not support the application of ultrasonography todetermine the presence or absence of tablets in an individual’s stomach afteran acute poisonings. We did show that, similar to other radiologictechniques (e.g., projection radiography), the absence of tablet visualizationdoes not exclude their presence. However, the ultrasonographers also foundseveral false positives. Thus, further studies are needed with larger numbersof both subjects and ultrasonographers to elucidate the potential benefit ofultrasonography in individuals who experience acute poisoning.

Key Words: ultrasonography, ultrasound, poisoning, overdose.(Pharmacotherapy 2011;31(3):273–276)

The frequency of acute poisonings remainshigh. In 2008, over 2 million poisonings werereported to poisons centers in the United States

alone.1 Many of these were intentional ingestionsin adults and unintentional ingestions inchildren. Despite the widespread occurrence of

PHARMACOTHERAPY Volume 31, Number 3, 2011

acute poisonings and the nearly infinite numberof drugs, only a few specific antidotes areavailable. Thus, techniques to decrease orprevent absorption are often employed in anattempt to reduce severe morbidity and, ideally,to prevent mortality from acute poisonings.Syrup of ipecac and gastric lavage have fallen outof favor since some of the risks associated withtheir use are considered to outweigh possiblebenefits.2–4 Other methods commonly employedinclude activated charcoal and whole bowelirrigation with polyethylene glycol–electrolytesolution.2, 5, 6 These techniques are thought to bemost beneficial soon after ingestion when thedrug is still in the stomach or just entering thesmall intestine.2, 5, 6 Unfortunately, no reliablemethod to visualize whether ingested tabletshave entered the stomach is currently available.

Projection radiography is sometimes used todetect the presence of an ingested drug, but it iseffective for only a small number of drugs, asmost are not radiopaque.7, 8 Therefore, a negativeprojection radiography finding does not excludethe presence of tablets. Computerized tomography(CT) has been shown both in experimentalmodels and case reports to identify some ingesteddrugs.9, 10 However, like projection radiography,many drugs are not detected. In addition, CTscans expose patients to a significant amount ofradiation, are expensive, and involve transport ofthe patient out of the emergency department.

Ultrasonography would appear to be anattractive option for identifying the presence ofingested drugs in individuals after an acutepoisoning. Ultrasound use in the emergencydepartment is common and has many applica-tions (e.g., gallbladder disease, trauma evaluation,pregnancy).11 Thus, many emergency physiciansare familiar with performing the ultrasoundtechnique, and some have received certificationin its use. Additional benefits of this techniqueinclude its portability, lack of radiation emission,and safe use in pregnant women. In addition,results can be immediately interpreted, enhancing

rapid response, which is paramount in a patientin an unstable condition, such as after an acutepoisoning. If drug presence could be consistentlyand correctly identified in a patient’s stomach,emergency personnel could more appropriatelyemploy gastrointestinal decontamination agents(activated charcoal, whole bowel irrigation).

Researchers used ultrasonography to visualizetablets in an in vitro artificial stomach model.12

In addition, they described visualization of anextended-release drug in a patient after acuteintentional ingestion.12 Thus, to further explorethe use of ultrasonography in this context, theobjective of our study was to determine whetherthis radiographic technique could correctly andconsistently identify the presence or absence ofingested drugs in human stomachs.

Methods

This prospective, ultrasonographer-blinded,pilot study was approved by the institution’sHuman Research Protection committee. Allsubjects provided written informed consentbefore participating. Subjects had to be healthy,at least 18 years old, and could not have ingestedany drugs within 24 hours of study enrollment.Exclusion criteria were pregnancy; known hyper-sensitivity or allergy to acetaminophen, aspirin,nonsteroidal antiinflammatory drugs, or multi-vitamins; and history of any chronic medicaldiseases (e.g., diabetes mellitus, peptic ulcerdisease, bleeding disorder, kidney or liverdisease). For women of childbearing age, a urinepregnancy test was performed before the study toensure the subject was not pregnant.

Subjects were randomized to one of two groups.The first group was assigned to ingest 500 ml ofwater alone (control group). The second groupwas assigned to ingest four acetaminophen 500-mg tablets (Perrigo, Allegan, MI), two multi-vitamin tablets (US Nutrition, San Francisco,CA), and four ibuprofen 200-mg tablets(Perrigo), all with 500 ml of water. Subjects inboth groups had to fast for at least 12 hoursbefore study participation.

Once subjects were enrolled, their age, weight,sex, and race-ethnicity were recorded. Eachsubject then underwent ultrasonography 2 timeswithin 30 minutes of ingesting the study tabletswith water or ingesting water alone. All subjectswere examined in the supine position. Thetransducer position was adjusted as needed tooptimize the image and allow visualization of thestomach. The same ultrasound machine was

274

From the Section of Toxicology, Department ofEmergency Medicine, University of Southern California, LosAngeles, California (Dr. Nordt); and the Department ofEmergency Medicine (Drs. Campbell, Medak, Tomaszweski,and Clark) and Division of Medical Toxicology (Drs.Tomaszweski and Clark), University of California–SanDiego, San Diego, California.

For reprints, visit http://www.atypon-link.com/PPI/loi/phco.For questions or comments, contact Sean Patrick Nordt,M.D., Pharm.D., Los Angeles County–USC Medical Center,1200 North State Street, Unit 1, Room 1011, Los Angeles,CA 90033; e-mail: spnordt@hotmail.com.

ULTRASONOGRAPHY FOR VISUALIZATION OF INGESTED TABLETS Nordt et al

used for all subjects to limit any potentialdifferences in the images obtained. Ultrasoundimages were obtained using a GE Logiq-Eportable ultrasound system with a phased-arraytransducer (GE Healthcare, Waukesha, WI).

Two board-certified emergency physicians withextensive ultrasound experience (both certifiedby their institution, one a registered diagnosticmedical sonographer) performed the ultrasono-graphy on each subject. The physicians evaluatedeach subject individually in isolation of the otherphysician so as not to influence each other’sinterpretations. The physicians were blinded towhich group the subjects were randomized. Thephysicians circled “yes” or “no” on a paper formto indicate their finding of presence or absence oftablets, respectively. The form contained nopatient identifiers other than a unique randomlygenerated letter and number combination, forwhich only the primary investigator had therandomization code.

Descriptive statistics were used for demographicdata. A false-negative response was defined astablets actually present in the stomach but notidentified by the ultrasonographer. A false-positive response was defined as no drugs presentin the stomach but incorrectly identified aspresent by the ultrasonographer. To assess inter-observer agreement between the two emergencyphysicians performing the ultrasonography, theCohen � coefficient was calculated.13 All datawere analyzed with Microsoft Office ExcelProfessional 2003 (Microsoft Corp., Redmond,WA).

Results

Fifteen subjects participated in the study; 12were male, three were female. The mean age was35.5 years (range 21–53 yrs), and mean weightwas 75.6 kg (range 50–97.7 kg). Nine subjects(60%) were randomly assigned to ingest thetablets, and six subjects (40%) were randomizedto the control group.

One of the ultrasonographers was correct indetermining whether or not tablets were visiblein the stomach for 14 (93%) of the 15 subjects.The other ultrasonographer, however, was onlycorrect for 6 subjects (40%). Six false negativesand four false positives were found. No inter-observer agreement was found between the twoultrasonographers (Cohen � coefficient –0.25[95% confidence interval –0.7–0.2]).

None of the subjects experienced any adverseevents during the study.

Discussion

Only a few specific antidotes are available totreat acute poisonings14; thus, techniques such asactivated charcoal and whole bowel irrigation arecritical in severe or potentially fatal poisonings toprevent or minimize the absorption of ingestedtoxins. Activated charcoal is generally only usedwithin 1 hour of an acute ingestion.5 However,the presence of tablets visualized in the stomachby ultrasonography beyond 1 hour could supportextended use of activated charcoal. Certaindrugs ingested during an acute poisoning maydecrease stomach emptying time and gastro-intestinal transit time. As such, “late” adminis-tration of gastrointestinal decontaminationagents may be beneficial in these situations.

Whole bowel irrigation is generally employedafter acute poisonings of certain agents such asiron tablets, sustained-release preparations ofcalcium channel blockers and lithium, enteric-coated drugs, and illicit drug packets.6 Confirmingthe presence of these drugs in the stomach byultrasonography could assist in the decision toinitiate this technique.

However, none of these decontaminationtechniques has been shown to decrease mortalityexcept in one study in which multidose activatedcharcoal administration after yellow oleanderpoisoning improved survival.15 A subsequentsimilar study failed to demonstrate this benefit.16

We specifically used three different drugs sothat we had variability in tablet size, density, andhardness, all which may provide differentacoustic signals. It is unclear whether very smalltablets could be identified. However, for safetyreasons, we used only 10 tablets/subject. Acutepoisonings usually consist of much largeringestions (e.g., 100 tablets). If such a largequantity of even smaller tablets was ingested, itmay be easier to visualize them based on sheernumber alone. We also chose to administer over-the-counter drugs at twice their therapeuticdoses in order to maintain a favorable safetyprofile.

Limitations

There are several limitations to our study.First, we used only two emergency physicians toperform the ultrasound studies, both of whomhave extensive ultrasound experience. Ourreasoning was that these individuals could moreeasily identify the presence of ingested tabletscompared with others who are less facile with

275

PHARMACOTHERAPY Volume 31, Number 3, 2011

ultrasonography. Still, consistent agreementbetween their evaluations was not found. It isunknown whether the use of more ultrasono-graphers would have demonstrated betteragreement.

In addition, our subjects had fasted for at least12 hours before participating in the study. It isnot clear what effect the presence of food in thestomach would have had on the ability tovisualize tablets correctly. Also, we administeredonly 10 over-the-counter drug tablets for safetyreasons. It is possible that if more tablets wereingested as in a true acute poisoning, they couldhave been more easily visualized. Finally,subjects took their tablets with water to act as anacoustic medium in the partially gas-filledstomach. It took 15–30 minutes after ingestionfor both physicians to perform their individualultrasound studies. This potentially allowedsufficient time for gastric absorption andemptying of the water, possibly resulting in lossof acoustic medium for the ultrasound. It ispossible that if subjects had ingested anadditional equal volume of water immediatelybefore each ultrasound study, the acoustic signalfrom the tablets could have been improved.

Conclusion

Our results do not support the routine orwidespread application of ultrasonography afteracute poisonings. We did show that, similar toother radiologic techniques, the absence ofvisualization of tablets does not exclude theirpresence. However, the ultrasonographers alsofound several false positives. Fortunately, theadministration of routine decontamination agents(activated charcoal, whole bowel irrigation) iswell tolerated and safe in the vast majority ofpatients. Nonetheless, we do not support anytreatment decisions being made based on bedsideultrasound use in patients after acute poisonings.Further studies are needed with larger numbers

of both subjects and ultrasonographers toelucidate the potential benefit of ultrasonographyin this setting.

References1. Bronstein AC, Spyker DA, Cantilena LR, et al., 2008 annual

report of the American Association of Poison Control Centers’national poison data system (NPDS): 25th annual report. ClinToxicol 2009;47:911–1084.

2. Manoguerra AS. Gastrointestinal decontamination afterpoisoning. Where is the science? Crit Care Clin 1997;13:709–25.

3. Krenzelok EP, McGuigan M, Lheur P, for the AmericanAcademy of Clinical Toxicology and the EuropeanAssociation of Poisons Centres and Clinical Toxicologists.Position statement: ipecac syrup. Clin Toxicol 1997;35:699–709.

4. American Academy of Clinical Toxicology and EuropeanAssociation of Poisons Centres and Clinical Toxicologists.Position statement: gastric lavage. Clin Toxicol 1997;35;711–19.

5. Chyka PA, Seger D. Position paper: single-dose activatedcharcoal. Clin Toxicol 2005;43:61–87.

6. Tenenbein M, for the American Academy of ClinicalToxicology and the European Association of Poisons Centresand Clinical Toxicologists. Position statement: whole bowelirrigation. Clin Toxicol 1997;35:753–62.

7. O’Brien RP, McGeehan PA, Helmeczi AW, et al. Detectability ofdrug tablets and capsules by plain radiography. Am J EmergMed 1986;4:302–12.

8. Tillman DJ, Ruggles DL, Leiken JB. Radiopacity study ofextended-release formulations using digitalized radiography.Am J Emerg Med 1994;12:310–14.

9. Savitt DL, Hawkins HH, Robers JR. The radiopacity ofingested medications. Ann Emerg Med 1987;16:331–9.

10. Judge BS, Hoyle JD. Incidental discovery of radiopaque pills onabdominal CT in a patient with abdominal pain. Clin Toxicol2008;46:574–5.

11. American College of Emergency Physicians. ACEP emergencyultrasound guidelines-2001. Ann Emerg Med 2001;38:470–81.

12. Amitai Y, Silver B, Leiken JB, Frischer H. Visualization ofingested medications in stomach by ultrasound. Am J EmergMed 1992;10:18–23.

13. Cohen J. A coefficient agreement for nominal scales. EducPsycholog Measurement 1060;20:37–46.

14. Dart RC, Borrow SW, Caravati EM, et al. Expert consensusguidelines for stocking of antidotes in hospitals that provideemergency care. Ann Emerg Med 2009;54:386–94.

15. de Silva HA, Fonseka MM, Pathmeswaran A, et al. Multiple-dose activated charcoal for treatment of yellow oleanderpoisoning: a single-blind, randomized, placebo-controlled trial.Lancet 2003;361:1935–8.

16. Eddleston M, Juszczak E, Buckley NA, et al. Multiple-doseactivated charcoal in acute self-poisoning: a randomizedcontrolled trial. Lancet 2008;371:579–87.

276