Objective. To determine whether parental errors in dosing liquid medication can be decreased through education.
Design. Randomized convenience sample stratified to three study groups.
Setting. General pediatric clinic, largely indigent and Latino.
Patients. A total of 45 English-speaking and 45 Spanish-speaking children diagnosed with otitis media and treated with an antibiotic suspension.
Intervention. Group 1 patients received the prescription and verbal instructions. Group 2 patients received the prescription and a syringe, then the correct dose was demonstrated. Group 3 patients received the prescription, a syringe with a line marked at the correct dose, and a demonstration. After returning from the pharmacy, parents administered the medication under observation. Parents in group 1 used a dispensing device similar to that planned for home use. The other groups used the syringe. After observation but before discharge, everyone received a syringe with a line marked at the correct dose. Patients were seen again at ∼1 month, and parents demonstrated how much medication they had administered.
Main Outcome Measure. Percent of parents who administered the correct dose.
Results. Patients in group 1 received between 32% and 147% of the correct dose, with only 11 of 30 (37%) receiving the correct dose (±0.2 mL). In group 2, 25 of 30 (83%) parents administered the correct dose, and in group 3, 30 of 30 (100%) gave the correct dose. Simultaneous logistic regression indicated that accuracy of dosage differed across instructional groups and language. At follow-up, 23 of 26 parents demonstrated the correct dose.
Conclusion. Education can decrease medication dosing errors made by both Spanish-speaking and English-speaking parents. Effectiveness was also shown at follow-up.
Most preschool-age children who are given medications receive them in a suspension. This can lead to many errors; children may spit or vomit the medication and some preparations expire within 2 weeks, must be resuspended before each dose, and may need refrigeration to maintain effectiveness.1 Perhaps the most common errors occur, however, when measuring the medication. For example, Gribetz and Cronley reported that parents underdosed acetaminophen, because they used the dropper provided with the infant preparation when dosing the children's suspension.2 Many physicians prescribe antibiotics using a teaspoon as the unit measure. Several studies have documented the inaccuracy of this method, especially when families use household teaspoons instead of a measuring teaspoon (the volume of teaspoons ranges from 2 to 9 mL).3-6 Research has suggested that parents may be confused about differences among teaspoons, tablespoons, and dose cups.2,7,8 Problems can also result from spillage and medication left in or on the measurement device rather than administered to the child.3,9 We chose to study the ability of parents to administer an antibiotic suspension prescribed for otitis media both because of the frequent diagnosis of this condition and because medication-administration skills could be reassessed easily during a follow-up visit. This study was only intended to determine whether the prescribed dose could be administered correctly by parents. We did not attempt to determine what dose was medically appropriate, nor did we study how accurately a dose must be measured to treat otitis media effectively. Parents were instructed to give the dose as written by the examining physician.
We intended to measure the frequency of dosing errors under three conditions: 1) using utensils that would commonly be used for measurement in the home, 2) using a syringe with verbal instructions, and 3) using a syringe marked at the correct dose by the researcher. We hypothesized that using a syringe with clearly marked units and a line drawn at the correct dose would significantly decrease measurement errors by ≥30%. A syringe was chosen because of its clearly marked units, accuracy, and utility with small volumes. We also intended to test the relative efficacy of this approach in Spanish-speaking and English-speaking families. The null hypothesis was that there would be no difference in measuring errors across the different instructional conditions or between Spanish-speaking and English-speaking families.
Children <4 years of age diagnosed with otitis media and placed on an antibiotic suspension were candidates to participate in part 1 of the study. A follow-up examination of the otitis media within 6 weeks of the original diagnosis allowed participation in part 2 of the study. Patients received care in the Comprehensive Health Care Pediatric Clinic at the Maricopa Medical Center in Phoenix, AZ. The clinic primarily serves families on the Arizona Health Care Cost Containment System, a managed-care Medicaid program. Some patients, however, have no insurance or have traditional fee-for-service coverage. The clinic also serves a large Latino population that resides in central Phoenix. To participate in the study, the patient's prescription had to be filled at the clinic pharmacy; if the family chose to fill the prescription elsewhere, the patient was not included in the study. Because of their health plans, most patients used the clinic pharmacy. The sample was convenience-based; all children who were eligible participated whenever the researcher was available in the clinic. Patients were stratified into three different categories. Groups 1 to 3 were filled randomly in this order. Patients who spoke Spanish only were given verbal instructions in Spanish. The researcher spoke medical Spanish; however, all nursing assistants in the clinic were bilingual, and there were two translators available if necessary. The researcher never acted as the examining physician. The examiners included residents, medical students, and attending physicians. The examiner notified the researcher whenever the diagnosis of otitis media was made and an antibiotic suspension indicated. The researcher then assigned the patient to group 1, 2, or 3, and gave the family the prescription and instructions as outlined below for the specific group.
Informed consent was not obtained, because the study was a quality-control audit to determine the number of patients using measuring devices correctly. The treatment was within standard of care for otitis media.
Parents of group 1 patients were given the prescription by the researcher, and written directions from the examining physician were read to them. (For example, “Take 1 teaspoon by mouth three times a day for 10 days.”) Group 2 participants received the prescription, verbal instructions, and a syringe (Exacta medicine dispenser, 10 mL, Baxa Corporation, Englewood, CO). The researcher demonstrated the dose by pulling back the plunger to the correct measure. Parents gave verbal confirmation that they understood the dosage to be given. Group 3 families received the same instructions as group 2, but the syringe also was permanently marked at the correct dose with a “T”. (The “T” was a vertical line drawn from 0 to the correct dose; the horizontal line was drawn around the syringe at the correct dose.)
These three groups also were divided based on language. Half of the patients in each of the three groups spoke Spanish only and half spoke English. If participants had asked for more information than provided for their category, this would have been noted and the information given at the end of the study. However, no one asked for more information.
The participants then were sent to the pharmacy to pick up the medication and asked to return to the clinic with the medicine and the child so that a dose could be administered before leaving the building. Average return time from the pharmacy was 30 minutes (range: 10 to 60 minutes). The participants in group 1 were asked how they planned to administer the medicine at home. The researcher had several measuring devices available including household teaspoons and tablespoons, baby spoons, plastic spoons, measuring spoons, dosing spoons, 3-teaspoon dosing cups, 0.8 and 5 mL droppers, and syringes. The parent then used the item most similar to that planned for home use to measure the dose of antibiotic. The researcher measured the dose to the nearest 0.2 mL, using a syringe to determine the accuracy of the dose. If the dose was correct (±0.2 mL), parents were asked if they wanted to use the syringe to administer the medicine. If so, the syringe was labeled at the correct line with a “T”. Future use of household teaspoons was discouraged. If the parent measured an incorrect dose, the dose was corrected and demonstrated again. All parents were given a syringe with the correct dose marked before leaving the clinic.
Parents in groups 2 and 3 also were asked to give their child a dose of medication using the syringe they had received previously. If parents had refused to use the syringe, they would have followed the protocol for group 1 (none refused). The dose drawn by the parent was measured to the nearest 0.2 mL by the researcher. If the dose was inaccurate by >0.2 mL, the parent was instructed again. The researcher also drew a “T” on all group 2 syringes after measuring the dose. The appropriate line was drawn previously for those in group 3.
Parents were asked to return for follow-up care of the ear infections in ∼1 month, and appointments were made for them. The researcher used a flow chart to document the patient's name, identification number, date, dose prescribed, dose given, device used, language, ethnicity, and date for follow-up.
During the study, the clinic pharmacy continued its usual practice of providing a dosing spoon with all liquid medications and explaining to parents the written instructions on the bottle. The spoons were molded plastic with raised numbers and markings. It was standard practice to provide a syringe for prescriptions written in milliliters or for very small volumes. The pharmacist did not demonstrate how to use these measuring devices. Arizona law requires pharmacists to provide instructions; therefore, this policy continued throughout the study.
Part 2 of the study focused on the long-term effectiveness of the education. Parents of patients entered in the study should have left the clinic able to use a syringe to the nearest 0.2 mL. When patients returned for follow-up, parents were asked again how much medication was administered, how many times a day, and for how many days. They also were asked which device they used and to demonstrate the amount administered. The demonstration syringe did not have the “T” marked on it. If parents said that they used the syringe and drew the medication up to the line that had been drawn at the last visit, another line was drawn to assist them.
Simultaneous logistic regression was used to assess the differences in accuracy across instructional language and treatment groups, concurrently. Follow-up Fisher's exact tests were used to evaluate specific differences identified in the initial analysis.
A total of 90 patients entered the study; 45 spoke English and 45 spoke Spanish. They were stratified randomly to three study groups, with 30 participants per group. Among the English-speaking participants, there were 30 Latinos, 9 Euro-Americans, 5 African-Americans, and 1 Native American. All of the Spanish-speaking families were classified as Latino. In Phoenix, ∼90% of Latinos are Mexican-American. Of the parents present, 97% were mothers.
Ten patients who were potentially eligible were excluded from the study. Six patients did not return from pharmacy, three patients chose to use another pharmacy, and one patient was accompanied by relatives who did not live in the home and who would not be available for follow-up.
The dose of antibiotic prescribed is shown in Figure1. The majority of patients (68%) were prescribed 1 teaspoon, 21% received 0.75 teaspoon, and the remaining 11% were prescribed various doses.
Group 1 parents gave 32% to 147% of the dose prescribed, with 11 of 30 (37%) dosing correctly. Mean absolute percent error was 22%. In group 2, the dose given ranged between 20% and 152% of the correct dose, with 25 of 30 (83%) dosing correctly. Mean absolute percent error in this group was 7%. Finally in group 3, all 30 parents, or 100%, gave the correct dose (Figure 2).
To assess differential accuracy of dosing among groups, binary outcome data were used; parents either gave an accurate dose (within 0.2 mL) or did not. Simultaneous logistic regression, with group coded as a categorical variable, indicated group (P = .01) and language (P = .03) differences. English-speaking participants were accurate in 86.7% of observations and Spanish-speaking in 71.1% of observations. The interaction of group by language (ie, whether the impact of instruction differed across language groups) was not significant (P = .33). Follow-up analyses on instructional group using two-tailed exact tests showed that the accuracy of group 1 differed significantly from that of group 2 (P = .005) and group 3 (P = .0001); however, groups 2 and 3 did not differ (P = .11).
A total of 35 patients returned for follow-up, but only 26 were seen by the researcher (16 Spanish-speaking and 10 English-speaking). Mean days at follow-up was 28 (range: 16 to 41 days). Those returning were divided equally among the three study groups. Dosing accuracy of participants who returned for part 2 (19.1% incorrect in part 1) did not differ significantly from those who failed to return (27% incorrect in part 1) (P = .41). All parents said they had used the syringe to administer the medication to their child. Of the 26, 23 demonstrated the correct dose on the syringe. Of the Spanish-speaking families, three requested that the line be drawn on the syringe and then demonstrated the correct dose. All families knew the number of times per day the medicine was to be administered. However, two families stated that they administered the medication for 8 days only, and one family used the medicine for 14 days. The remaining parents stated they used the medication for 10 days as directed.
This study has shown that parental education can be very effective in eliminating medication dosing errors. Fully 100% of English-speaking and Spanish-speaking families dosed the medication correctly when given instructions and a syringe with a line marked at the prescribed dose. Families must be educated to ensure the best possible outcome for children. It is possible that some treatment failures are attributable to incorrect dosing of medication rather than to incorrect choice of medication.10-12 Parents must know that health care professionals want to help educate families about diseases and medications, not just make diagnoses and write prescriptions. Better communication between medical staff and family is essential.13-17 Parents may want to do what is best for their child but feel overwhelmed in the medical environment.11 We need to empower them with information. Office staff should ask all families whether they were satisfied with their visit and understand the treatment plan.
The role of the pharmacist requires clarification in each practice. Many pharmacists do not provide information about medications.5,17 A phone survey of 29 pharmacies located near the clinic showed that 31% of pharmacies supplied devices with suspensions, 45% did so occasionally (varied with age, type of medication, and whether requested), and 24% did not provide any device. We should insist on uniform labeling and provide labeled measuring devices with liquid medications.5,8,11,15,18Managed care plans should determine whether the pharmacy or clinic will provide the measuring device. A syringe provides an accurate measure, is easy to control, causes minimal spillage, may be used in viscous fluids, and is useful for small infants because it initiates sucking compared with use of a dosing spoon.9 Some problems with the syringe are cost, dexterity required, choking if delivered too fast, and possible use as an intravenous device.9,19 Also, several parents in this study drew air bubbles in the syringe.
In the study by Mattar, Markello, and Yaffe, when no dispensing device was given, 71% of parents used a teaspoon.5 This practice can be eliminated through better education and by providing labeled measuring devices. Although accuracy to the nearest 0.2 mL is not required for all medications, parents are able to dose medication correctly when given an accurate measuring tool and education. The type of device recommended for dosing should be determined by each clinician. Considerations should include the degree of accuracy required, cost, and the age of the patient. Health care professionals must take a thorough history of how the medication was dosed before assuming that treatment failure was attributable to the medication prescribed. Additional studies should be conducted to determine whether after using the method of education described, compliance in taking medication can be increased.
- Received September 16, 1996.
- Accepted January 29, 1997.
Reprint requests to (S.R.M.) Department of Pediatrics, Maricopa Medical Center, 2601 E Roosevelt, Phoenix, AZ 85008.
- Schentag JJ
- Gribetz G,
- Cronley SA
- McKenzie MW
- Arnhold RG,
- Abebonojo FO,
- Callas ER,
- Callas J,
- Carte E,
- Stein RC
- Hyam E,
- Brawer M,
- Herman J,
- Zvieli S
- Jonville AE,
- Autret E,
- Bavoux F,
- Bertrand PP,
- Barbier P,
- Gauchez AM
- ↵Kimminau MD. New findings on oral dosing. Am Pharm. 1980;MS20:39–41
- Mattar ME,
- Yaffe SJ
- Faigel HC
- Mattar ME,
- Markello J,
- Yaffe SJ
- Committee on Drugs
- Ellison RS,
- Altemeier WA
- Copyright © 1997 American Academy of Pediatrics