Article Text


Accurately administering oral medication to children isn't child's play
  1. Victoria L Beckett1,
  2. Luke D Tyson1,
  3. Daniel Carroll2,
  4. Nigel M Gooding3,
  5. A W Kelsall2
  1. 1School of Clinical Medicine, University of Cambridge, Addenbrooke's Hospital, Cambridge, UK
  2. 2Department of Paediatrics, Addenbrooke's Hospital, Cambridge, UK
  3. 3Pharmacy Department, Addenbrooke's Hospital, Cambridge, UK
  1. Correspondence to Dr Victoria L Beckett, School of Clinical Medicine, University of Cambridge, Addenbrooke's Hospital, Box 111, Hills Road, Cambridge CB2 0SP, UK; victoria.beckett{at}


Objective Parents administer oral medications with various measuring devices including metal teaspoons, calibrated spoons and oral syringes. We aimed to determine which was the most accurate.

Design Self-controlled, non-randomised, experimental study.

Setting Caregivers attending paediatric outpatient clinics.

Methodology Caregivers measured 5 ml of 120 mg/5 ml paracetamol suspension using a 5.0 ml metal teaspoon, 5.0 ml calibrated spoon and 5.0 ml oral syringe. Samples were weighed and converted to mls.

Main outcome measures The mean volume and variance of volumes were measured for each device.

Results We recruited 277 caregivers (98% parents). Volumes measured ranged from 0.83–6.52 ml. Accuracy did not vary with caregivers’ age, gender, instrument preference, number and age of children. The mean volumes measured with the oral syringe (95% CI 5.09 to 5.17 ml) and metal spoon (95% CI 3.90 to 4.08 ml) were significantly different to the desired 5 ml volume (p<0.0001), dissimilar to the mean volume measured using the calibrated spoon (95% CI 4.91 to 5.09 ml, p=0.99). The variance of volumes measured with the oral syringe (SD 0.348 ml) was significantly smaller (p<0.0001) than that measured using a calibrated spoon (SD 0.762 ml) or metal spoon (SD 0.749 ml).

Conclusions The calibrated spoon was the most accurate producing a mean volume of 5 ml, while the oral syringe had the smallest variance. The increased variability of calibrated or metal spoons may result in under or overdosing especially when administering drugs with a narrow therapeutic window. Health care professionals must make a case-by-case decision regarding which device is preferable depending on the medication in question. Parental education could improve measuring accuracy.

Statistics from

What is already known on this topic

  • Parents routinely administer sub-therapeutic doses of medication.

  • The use of a metal spoon to measure and administer liquid medication is significantly inaccurate due to inconsistencies in spoon volume.

What this study adds

  • Parents are most accurate when measuring and administering oral medication to children using a calibrated spoon, however the oral syringe produces the smallest variance.

  • Health care professionals should take this into consideration when providing parents with measuring devices, which may well depend on the medication in question.

  • Parents should be educated on correct use of the oral syringe to further improve dosing accuracy.


The administration of oral medication to children by caregivers is extremely topical.1 ,2 The potential for dosing errors is greater in children than adults3–7 because paediatric doses are dependent upon age, weight and body surface area.3 ,4 ,8 Several studies estimate that 62–80% of parents routinely administer sub-therapeutic (<10 mg/kg) doses of paracetamol solution to their children.3 ,7 ,9 ,10 Parental education regarding measuring drug doses is often limited, leading to dosing inaccuracy with potentially harmful consequences.11 Errors may also occur when parents administer their child's medication. Most preschool children receive oral medication as a suspension, which is routinely supplied with a calibrated measuring spoon or oral syringe to measure daily doses. When oral suspensions are not supplied with a measuring device, 75% of parents will select a kitchen spoon to administer children's medication.12

It is well established that measuring and administering oral medication using a metal kitchen spoon is inaccurate, principally due to the significant variability of teaspoon volumes which can range between 1.5 and 9 ml.3 ,12–15 However, few studies have evaluated the accuracy of other measuring devices and none have directly compared the 5 ml oral syringe with the 5 ml calibrated spoon. Sobhani et al12 determined that caregivers were more accurate measuring 5 ml paracetamol suspension using an oral syringe compared with a dosing cup (p<0.001). Yin et al16 reported some differences in the accuracy of calibrated spoons compared with oral syringes, but did not comment on significance.

We hypothesised that there could be a difference between the accuracy of the calibrated spoon and the oral syringe when used by caregivers. In this study, we aimed to determine which measuring device (metal teaspoon, calibrated dosing spoon or oral syringe) was the most consistently accurate when measuring 5.0 ml of standard paracetamol suspension, generating recommendations for future practice. For medications with a narrow therapeutic window such as carbamazepine or phenytoin, consistent under- or overdosing caused by a measuring inaccuracy could have serious consequences.

We additionally explored caregivers’ views and preferences for measuring devices used in the home environment and investigated whether caregiver's demographics influenced the accuracy of medication measurement and administration.


We recruited 277 caregivers (parents and carers) attending paediatric outpatient clinics at Addenbrooke's Hospital, Cambridge during September and October 2010. Caregivers attending clinic were asked to enrol prior to their appointment, and included if they consented. Five caregivers declined to participate due to time constraints. Study participants completed a questionnaire detailing the number and age of children living at home, regularity of and responsibility for administering children's medication, preference and perceived accuracy of paediatric medication devices. The latter two responses were tested both before and after measuring paracetamol. Caregivers were asked to measure 5 ml of paracetamol suspension (Orbis Pharmaceuticals, 120 mg/5 ml) using a 5 ml metal teaspoon, 5 ml calibrated spoon dispensed with the suspension and 5 ml oral syringe (Medicina oral/enteral 5.0 ml). Each instrument was weighed before and after measuring using 50 g pocket scales accurate to ±0.005 g. The volume of paracetamol measured was calculated using the conversion factor 1.09 g/ml (Orbis Pharmaceuticals). The order of device use was derived from a random number generator. One teaspoon and calibrated spoon were used throughout but syringes were changed for practical reasons. Spoons were thoroughly washed and dried to prevent contamination.

Single-sample t-tests compared the mean volume measured using each instrument to the desired volume of 5 ml, and paired t-tests compared the mean volumes measured with the different devices. To compare the variance of volumes measured with the three measuring devices we used Pitman's Test of Difference in Variance17 to ascertain the correlation. For all analyses two tailed p<0.05 was considered statistically significant. The 95% reference range of volumes measured with each device was used to extrapolate our data to examine the dosing variation that may be generated. Spearman's Rank Correlation (two tailed analysis) was employed to determine whether caregiver accuracy varied according to the age of the caregiver or child, where p<0.05 was deemed statistically significant. 95% CI were calculated for the volume of paracetamol measured depending on caregiver gender, number of children, frequency of medication administration or preference of measuring device. Overlapping CI signalled statistical insignificance. We conducted a preliminary experiment (table 1) to pilot the feasibility of the study and perform a statistical power calculation.

Table 1

Results from the preliminary experiment, n=21 caregivers

For every caregiver tested (n=21) we calculated the difference in the measured volume using each instrument and determined the SD of these differences. This was 0.63 ml comparing metal with calibrated spoons, 0.82 ml comparing the metal spoon with the oral syringe, and 0.76 ml comparing the oral syringe with the calibrated spoon.

Hence, we assumed the SD of the differences between each device to be 0.80 ml and used this figure to perform an 80% and 90% power calculation where p≤0.05 was significant. Using a two-tailed paired sample t-test a sample of 275 would have 80% power to detect a mean difference of 0.14 ml between two measurement devices, or 90% power to detect a difference of 0.16 ml.


Two hundred and eighty-two caregivers were approached to participate in this study, five declined and 277 caregivers aged 17–68 years (mean 36.3; 95% CI 35.3 to 37.2 years) were recruited. Almost all (n=272, 98%) were parents with children less than 18 years of age at home (mean age 7.94 years; 95% CI 7.46 to 8.42 years).The remainder were grandparents or step-parents. Most would choose an oral syringe (n=164, 59%) or calibrated spoon (n=146, 52%) to dispense medication. Caregivers were given the option to select more than one device. The majority (n=232, 84%) considered the syringe most accurate and easiest to use (n=195, 70%). Medication was mainly administered by mothers (n=261, 94%). Measuring accuracy did not vary with caregivers’ age, gender, number and age of children, frequency of medication administration or preference of measuring device.

The mean volumes measured for all devices are shown in table 2. Using the calibrated dosing spoon the mean volume was 5.00 ml (95% CI 4.91 to 5.09 ml). This was not significantly different to the 5.0 ml ideal volume (single-sample t-test, p=0.9931) but was significantly different to the mean volume measured with the metal spoon (paired t-test, p<0.0001) and oral syringe (paired t-test, p=0.0064).

Table 2

Mean volume, SD and range of volumes measured with each device n=277, ideal volume=5.0 ml

The mean volume measured with the oral syringe (5.13 ml; 95% CI 5.09 to 5.17 ml) was significantly greater than the ideal volume of 5 ml (single-sample t-test, p<0.0001) and significantly different to the mean volume measured using the calibrated spoon (paired t-test, p=0.0064) or metal spoon (paired t-test, p<0.0001).

The most inaccurate measuring device was the metal spoon, which generated a mean volume significantly less than 5 m (3.99 ml; 95% CI 3.90 to 4.08 ml, single-sample t-test, p<0.0001) with a wide measuring range of 0.83–6.51 ml. The mean volume measured using the metal spoon was significantly different to that of the oral syringe and calibrated spoon (paired t-test p<0.0001 in both instances).

The variance of volumes measured with the oral syringe was significantly smaller than the metal and calibrated spoons (calibrated spoon vs syringe, Pitman's test <0.0001, metal spoon vs syringe Pitman's test p<0.0001). The variance of the calibrated spoon did not differ significantly to that of the metal spoon (Pitman's test, p=0.5066).

To examine the clinical relevance, we calculated the range of paracetamol doses (in mg) that 95% of caregivers would dispense using an oral syringe, metal or calibrated spoon. This calculation was based upon the 95% reference range of volumes obtained with each measuring device.

Applying this to a clinical scenario, a 5-year-old child weighing 20 kg would require a 250 mg dose (5 ml) of paracetamol. Ninety five percent of caregivers would dispense 125–275 mg of paracetamol with a metal spoon, 174–326 mg using a calibrated spoon and 222–291 mg with an oral syringe. This dosing variation reflects the inconsistency (variance) of the volumes measured with each instrument. The range and consequently the variance of doses dispensed with the oral syringe are considerably smaller than both the calibrated and metal spoons.

This may prove particularly important for medications with a narrow therapeutic index such as carbamazepine. We extrapolated our results (the 95% reference ranges of volume measured) to carbamazepine oral suspension. A 20 kg child would require 5 mg/kg or 100 mg (5 m) of carbamazepine. Ninety five percent of caregivers would dispense a carbamazepine dose between 50–110 mg with a metal spoon, 70–130 mg using a calibrated spoon and 89–116 mg with an oral syringe. The range of doses measured using a metal or calibrated spoon correspond to a 12.0 kg weight error. With the oral syringe the range of measured doses generates a 5.6 kg error in the weight of the child, which although smaller, could still be significant given the child weighs only 20 kg. The calibrated spoon would have the greatest accuracy measuring a dose of 5 ml.


This study has provided information about caregivers’ accuracy when measuring liquid medication for children. In accordance with previous evidence7 ,18 our study suggests that demographic factors (including age, sex, number of children) do not affect caregivers’ ability to dispense medication. However, there is some evidence that women and more educated caregiver's more frequently measure an accurate dose, administering this at the correct time and frequency.18

This study confirms the inaccuracy of household teaspoons for measuring oral medication, a topic which has attracted considerable media interest.1–3 ,6 ,12 ,15 ,18–21 Within this study the wide range of volumes measured with the metal teaspoon (0.83–6.51 ml) supports previous literature quoting a 2.5–7.3 ml variation. Caregivers using the largest household teaspoons may administer almost three times as much medication than a caregiver using the smallest teaspoon.1 ,2 ,12 ,15

In our study most caregivers preferred a calibrated device to measure liquid medication (59% (n=164) selecting an oral syringe and 52% (n=146) choosing a calibrated spoon), which is higher than in previous series where 73% of caregivers primarily use a household teaspoon.13 ,18

The mean volume measured using the calibrated spoon (5.00 ml) was not significantly different to the 5 ml ideal volume. Previous literature has generated mixed conclusions concerning calibrated spoon accuracy. Ellerbeck et al22 found that 38% of caregivers (n=400) using calibrated spoons under dosed children by ≥30%; a finding which was restricted to children >1 year of age. On the contrary, Kairuz et al23 determined that calibrated spoons generated a mean overdose of 11.8%, however in this study caregivers measured only 3 ml of liquid suspension. Dissimilar to our results, Yin et al concluded that the mean dose measured with the calibrated spoon was greater than the syringe, but no statistical significance calculations were performed. Interestingly, the adjusted adjusted odds (OR) of making no or small (±20%) measurement errors was higher with the calibrated spoon than the syringe, whereas larger dosing errors (>40%) were more common with the syringe.

The mean volume measured with the oral syringe (5.13 ml) was significantly greater than the ideal 5 ml volume. Previous studies using similar methodology found a mean volume of 4.5–4.6 ml (SD 0.7 ml) using an oral syringe (Becton, Dickinson, and Co, Franklin Lakes, New Jersey, USA; 5 ml), and 4.7 ml (SD 0.7 ml) using an oral syringe with a bottle adaptor, yet statistical calculations were not performed. (Medi-Dose, Inc/EPS, Inc, Ivyland, Pennsylvania, USA).12 ,16

Based upon the mean volumes measured the calibrated spoon is the most accurate device. However, the variance of the volumes measured with the calibrated spoon is significantly higher than that of the oral syringe (p=0.0001). The variance of each instrument essentially describes the range of volumes caregivers measure with that instrument; the smaller the variance the more consistent the dosing.

In this study the variance of the oral syringe (0.348 ml) was considerably lower than previously quoted values (SD=0.7 ml),12 reflecting our larger population size. No prior studies have directly compared the mean volume and variance of the calibrated spoon versus the oral syringe using statistical calculations.

We recommend that caregivers should routinely measure and administer liquid medication using either a calibrated spoon or oral syringe. Health care professionals should make a case-by-case decision regarding the provision of measuring devices to caregivers, which will also depend upon the medication in question. For example, health care professionals might recommend an oral syringe when prescribing medication with a narrow therapeutic index, where higher dosing variability could have more serious consequences.

In light of our findings, pharmaceutical manufacturers of paediatric medications (such as paracetamol suspension) should routinely provide both calibrated spoons and oral syringes in primary and secondary care. The large variance of volumes measured within our study suggests the need for improved parental education regarding the use of different measuring instruments, which has been shown to improve the accuracy of even the most consistent devices.6 ,12 ,18 ,19 Parents are rarely trained to accurately dose and administer oral medication, hence the majority (47%) underdose children, while 15% overdose.7 McMahon et al11 proved that caregiver education could improve measurement accuracy (measuring a dose within 0.2 ml of the intended dose) using an oral syringe from 37% to 100% (p=0.0001).

There were a few limitations to our study. Firstly the sampling was opportunistic from multispeciality outpatient clinics and not random. Secondly the participants in our study measured only paracetamol suspension, although we have extrapolated our results to carbamazepine. We did not account for caregiver education or dose calculation errors. Previous work has suggested that these account for a significant proportion (26%) of paediatric dosing errors.24 Finally, we only calculated the volume parents measured with each device, which does not equate to the volume of drug that the child receives. Administration of medication can be difficult when children are not cooperative. In addition some parents may transfer medication between devices, choosing to measure with an oral syringe and dispense with a spoon which may result in further variation in volume.


With thanks to The Cambridge Centre for Applied Medical Statistics (CAMS) for statistical advice.


View Abstract


  • Contributors VB co-led the planning, design and implementation of the experiment, interpretation of statistical analyses, drafted, reviewed and revised this paper. LT co-led the planning, design and implementation of the experiment, performing statistical analyses of the data, drafted and revised this paper. DC participated in design, planning and directed execution of the experiment. NG provided the experimental equipment, participating in planning and designing the intervention. WK participated in study design, planning and implantation, co-directing execution of the experiment, reviewing and editing drafts of this paper.

  • Funding None.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing Anonymised data from this study can be requested from Victoria Beckett (

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.