Prevalence of exceeding maximum daily dose of paracetamol, and seasonal variations in cold‐flu season

Abstract

Aims

To estimate prevalence of excess intake of paracetamol and investigate seasonal variations therein.

Methods

Between 2011 and 2016, 14 481 US adults who used paracetamol in the preceding 30 days were sampled from national online panels and completed a detailed online daily diary of paracetamol medication use for 7 days. Respondents were not told that the study concerned paracetamol. Cold/flu season (CFS), identified using Google Trends data, was contrasted to off‐season in symptoms, use of paracetamol medications, and consumption exceeding 4 g (the recommended daily maximum).

Results

Overall, 6.3% [95% confidence interval: 5.9–6.7%] of users exceeded 4 g on at least one day; 3.7% [3.5–3.8%] of usage days exceeded 4 g. Cold/flu symptoms were more likely to be experienced and treated with paracetamol in CFS than off‐season. Paracetamol users were more likely to exceed 4 g during CFS (6.5% vs. 5.3%; odds ratio = 1.24, 1.04–1.48); days exceeding 4 g also increased (3.9% vs. 2.8%; odds ratio = 1.37, 1.11–1.69). This was not due to differences in characteristics of individuals using paracetamol in CFS, but primarily to increased use of over‐the‐counter combination medications designed to treat upper respiratory cold/flu symptoms (33.2% of usage days in CFS vs. 24.8% in off‐season; odds ratio = 1.58, 1.46–1.72). When such medications were omitted, there was no statistically significant seasonal variation in exceeding 4 g.

Conclusions

Paracetamol use and over‐dosing increases in CFS, primarily due to increased use of over‐the‐counter combinations treating upper respiratory cold/flu symptoms. Pharmacists should warn users to follow labelled dosing directions, especially during CFS.

What is Already Known about this Subject

• Paracetamol is an active ingredient in many prescription and over‐the‐counter medications.

• When taken in high doses, paracetamol can cause liver injury.

• Users sometimes exceed the maximum recommended dose of 4 g per day.

What this Study Adds

• Over the course of 7 days, 6.3% of paracetamol users exceed 4 g on at least one day.

• Such overdose occurred more often in cold‐flu season (6.5%) than in off‐season (5.3%), due to increased use of over‐the‐counter combinations treating upper‐respiratory cold/flu symptoms.

Introduction

Paracetamol is an active ingredient present in hundreds of over‐the‐counter (OTC) and prescription (Rx) medications indicated for pain and fever as well as in many multi‐ingredient OTC medications additionally indicated for symptoms associated with colds, flu, allergies and sleeplessness 1. It is considered safe when dosed as directed, but, in overdose, has been associated with liver injury 1, and with calls to poison control centres (estimated around 112 000 year^–1), emergency room visits (59 000/year) and hospitalizations (38 000/year) 2, 3, 4. Many paracetamol over‐doses are the result of deliberate self‐harm, but a substantial proportion are attributed to unintentional over‐dose; for example, 30% of paracetamol‐related emergency room visits are attributed to unintentional over‐dose. The problem of unintentional over‐dosing was of sufficient concern for the US Food and Drug Administration (FDA) to convene an Advisory Committee 5, 6 on the subject in 2009.

The current maximum recommended daily adult dose of paracetamol is 4 g 7. In most instances, paracetamol‐related liver injury only occurs at considerably higher doses, thus limiting ingestion to 4 g would prevent injuries. Reports of retrospective assessments with small, select samples, such as emergency room patients 8 or dental clinic patients 9, suggest that paracetamol users periodically exceed the recommended daily dose in those settings. However, detailed prospective data on patterns of self‐treatment with paracetamol in the general population have been lacking until recently. We previously reported results from 7‐day medication diaries collected from paracetamol users during a single month (August 2010), estimating that the 4‐g daily limit was exceeded by 4.5% of paracetamol users and on 3.1% of paracetamol usage days 10. Analyses identified personal characteristics (e.g. chronic pain and smoking) and attitudes (e.g. that label directions are just suggestions), and usage patterns (e.g. concomitant use of multiple paracetamol medications, including a mix of single‐ingredient and combination medications) associated with exceeding the limit 10, 11.

One limitation of those results was that the data were collected during a single month in the summer, and so may not have optimally estimated the likelihood of exceeding the recommended maximum dose. Moreover, these data would not have captured seasonal influences on paracetamol use. Since many OTC combination medications are intended to treat the upper respiratory symptoms of cold and flu (designated upper respiratory medications), it seems plausible that use of upper respiratory medications would rise during the cold‐flu season, perhaps leading to an increase in exceeding 4 g. The present analyses estimate the prevalence of overdosing and assess seasonal variations in overdosing.

Methods

The study used the data‐collection methods described in detail in our previous publications 10, 11, but this is a new, nonoverlapping sample, collected between 7 March 2011 and 31 March 2016. The study was ruled exempt by the Boston University Medical Campus Institutional Review Board.

Participants

Respondents were US volunteers aged ≥18 years sampled from multiple online research panels (Lightspeed Research, GMI, Survey Sampling International) who responded to emails inviting them to participate in the study, in return for compensation. Nothing in the invitations or any other communication indicated that the study concerned paracetamol. The selection of panel members to be invited was adjusted as needed to the distribution of the US population, based on region, age, income, sex and race/ethnicity. Participants were restricted to those who reported using an paracetamol medication in the past 30 days, to increase the proportion who were likely to use paracetamol during the subsequent 7‐day diary period.

In total, 980 415 invitations were sent (it is not known how many were received or viewed), 168 016 persons visited the enrolment website, 48 803 qualified and 18 689 completed 7 days of diaries and an exit survey. The median age of participants was 44 years; 61% were female and 24% nonwhite; 46% reported income below $50 000; and 82% had education beyond high school.

Data collection

After enrolling, respondents were prompted by email daily for 7 consecutive days to complete an online diary of medication use. Each day after 3 pm, they recorded medications they had taken in the past 24 h (they were also allowed to fill in the previous day's use if missed), indicating how much they had taken and when, using a grid of hours that began at the time they completed the previous diary. To identify medications, respondents selected from a comprehensive list of paracetamol‐containing medications organized in a hierarchical structure by indication (pain/fever, cold/flu/sinus, allergy), class (OTC vs. Rx), and brand or product name (the list grew over time from 384 to 423 products, as new products became available). Respondents were not told that the list was limited to paracetamol‐containing medications. To reduce the burden of finding their medications for each diary entry, participants selected from their personal list of medications on hand, to which they could add at any time. After the week of diary‐keeping, participants completed an exit survey that assessed demographic characteristics, medication knowledge, exposure to products and product messaging, attitudes and some medical history. The diary and survey were developed by the authors and extensively pretested, through both qualitative methods and collection of pilot quantitative data.

Definition of key variables

Analyses focused on days of use (midnight to midnight) when dosing exceeded 4 g. Each time respondents reported taking a medication, they reported the product and how many units (e.g. tablets) they took, which enabled us to compute the daily dose. Within OTC combinations, we distinguished the medications intended (based on names and labelling) to treat upper respiratory symptoms (upper respiratory medications) from those designed to relieve other symptoms. For each dosing occasion, respondents were also asked to indicate what symptom or condition they were treating, selecting a primary and (optionally) a secondary target symptom from a list that included pain, fever, cold/flu, sinus, allergy, sleep and other.

Seasonality

To estimate seasonal variations in colds and flu, we used Google Trends 12 to analyse searches performed across the continental USA during the data collection periods. Analyses of Google searches have been used by epidemiologists to quantify the ebb and flow of various disease states 13, 14, and have been validated as indicators of seasonal variations in flu 15, 16, 17. We also validated the Google data against data on weekly purchase of all OTC medications classified as cough/cold/sinus medications (not limited to paracetamol medications), obtained from IMS (Danbury, CT, USA), for the 242 weeks from January 2009 to August 2013. The Google search data correlated 0.93 with the IMS data. The weekly search data were used to identify periods as being in the cold/flu season (CFS) or in the off season (OFFS), and these periods were contrasted in subsequent analyses. The seasonality curves also identified transitional periods between CFS and OFFS, when cold/flu searches were initially rising or falling; these periods were omitted from the analysis, as data were sparse and variable, precluding clear interpretation. While dates varied from year to year, CFS lasted roughly from September/October to March/April, while OFFS lasted from June to late August.

Analysis

We estimated the percentage of paracetamol users who exceeded 4 g or 8 g on at least 1 day and the percentage of usage days on which 4 g or 8 g was exceeded. The number of invitations per day varied over the course of the study; to account for this, and any confounding with seasonality, we weighted the data to even out the invitation rate. As reported below, this did not materially affect the outcome.

Analyses examining dosing occasions or days accounted for the nesting of observations within persons by using generalized estimating equations 18 with the logit link function and an exchangeable covariance structure for dichotomous variables, and mixed linear models with compound symmetry for continuous variables.

To assess whether seasonal differences in the profile of persons using paracetamol might account for any differences in the likelihood of exceeding 4 g, a multivariate model tested whether the variables that previous analyses 11 associated with exceeding the limit could differentiate CFS from OFFS users. This was modelled by logistic regression with season as the dependent variable and subject characteristics, assessed in the exit survey, as predictors (education, ongoing or chronic pain, daily smoking, 10 or more past‐year medical visits, SF‐12 physical component score 19, and attitude scores indicating "I know what dose is right for me," and "I start at the lowest dose," which were derived by factor analysis of medication attitudes) 11.

Results

Analytical samples

Overall, 18 689 individuals completed 7 days of diaries and the exit survey. Of these, 14 481 reported paracetamol use on 59 220 usage days (out of 101 367 diary days), comprising 122 854 dosing occasions; this was the dataset used in analyses estimating the overall probability of exceeding 4 g. Of the 14 481 users, 7991 (55.2%) reported using paracetamol on 4 or more diary days.

Overall rates of paracetamol over‐dosing

As shown in Table 1, across all seasons (including transitional periods), 6.3% (95% confidence interval: 5.9–6.7%) of users exceeded 4 g on one or more days, and on 3.7% (3.5–3.8%) of usage days; 8 g was exceeded by 1.2% (1.0–1.4%) of users, on 0.7% (0.6–0.7%) of usage days. Very few users [1.4% (1.2–1.6%)] exceeded 4 g on 4 or more diary days and only 0.2% (0.2–0.3%) exceeded 8 g 4 or more diary days. Estimates based on data weighted to account for variations in the number of invitations issued were nearly identical to the unweighted estimates for both 4 g (person: 6.4%; day: 3.7%) and 8 g (person: 1.2%; day: 0.6%); unweighted data were used for remaining analyses.

Table 1.

Percentage of paracetamol users and paracetamol usage days exceeding 4 and 8 g

	Percentage of users who exceeded a dose threshold on at least one day		Percentage of usage days on which a dose threshold was exceeded
	Estimate	95% CI	Estimate	95% CI
Exceeded 4 g a in a day	6.3%	5.9–6.7%	3.7%	3.5–3.8%
Exceeded 8 g in a day	1.2%	1.0–1.4%	0.7%	0.6–0.7%

Note: Unweighted data. Estimates based on data weighted to account for variations in recruitment rates produced nearly identical results

CI, confidence interval

^a4 g is the labelled maximum daily dose

Seasonal variations in symptoms and treatment

Participants' daily reports of symptoms they experienced (Appendix S1) showed the expected seasonal variation: cold/flu symptoms, fever and sinus symptoms were reported more often in CFS. The proportion of paracetamol dosing occasions intended to treat cold/flu, fever and sinus symptoms also increased in CFS (Appendix S2).

Seasonal variations in use and excess dosing of paracetamol

Analyses of seasonal patterns were based on data from CFS (9255 users, 64 785 diary days, 38 136 usage days, 80 229 dosing occasions) and OFFS (3271 users, 22 897 diary days, 13 121 usage days, 25 959 dosing occasions). The number of days of paracetamol use was very slightly higher in CFS (4.12 usage days per user) compared to OFFS (4.01; P = 0.014).

As shown in Table 2, on paracetamol usage days, the total dose was more likely to exceed 4 g in CFS than in OFFS, both when considering individual days and when considering the likelihood that individuals had at least 1 day exceeding 4 g. On a relative basis, compared to OFFS, the odds of a user exceeding 4 g on at least 1 day were 24% higher during CFS, and the odds of exceeding 4 g on an paracetamol usage day increased almost 40%. Similarly, in CFS, the odds of exceeding 4 g on 4 or more diary days are increased by 82% [1.5% vs. 0.9%; odds ratio (OR): 1.82, 1.21–2.73]. Numerically, there was a similar seasonal effect on exceeding 8 g (Table 2), but the effect was not statistically significant, likely due to the very low rates. There was no increase in exceeding 8 g on 4 or more days in CFS [0.2% in both CFS and OFFS, OR: 0.93 (0.41–2.10)].

Table 2.

Probability of acetaminophen use exceeding 4 or 8 grams according to season

	Persons			Days
	Off season	Cold/Flu season	OR (95% CI)	Off season	Cold/Flu season	OR (95% CI)
Exceeding 4 grams
All medications	5.3	6.5	1.24 (1.04–1.48)	2.8	3.9	1.37 (1.11–1.69)
Excluding UR medications	3.9	4.3	1.11 (0.90–1.36)	2.1	2.4	1.11 (0.87–1.43)
Exceeding 8 grams
All medications	0.9	1.3	1.41 (0.94–2.10)	0.6	0.7	1.30 (0.78–2.18)

Note: Prevalence for persons is based on exceeding 4 or 8 grams on at least one day, among persons who used acetaminophen during the 7‐day diary period; for days, it is the percentage of acetaminophen usage days in which 4 or 8 grams was exceeded. Based on data from March 2011 to March 2016.

Do user characteristics vary by season?

One possible explanation for seasonal variations in exceeding 4 g is that the individuals who use paracetamol during CFS differ from those using in OFFS in ways that increase their likelihood of exceeding 4 g. To test this, we compared users in CFS and OFFS on characteristics previously shown to be associated with exceeding 4 g 11. In a multivariate model, users of paracetamol in CFS did not differ significantly from users during OFFS in any of these characteristics (Appendix S3).

Seasonal variations in patterns of paracetamol use

The types of paracetamol medications used varied by season (Table 3). The proportion of days on which OTC combination medications were used – either alone or along with single‐ingredient medications – increased (ORs 1.2–1.3). Among the OTC combinations, only those classified as upper respiratory medications increased in CFS (OR = 1.6), while the use of allergy medications declined (OR = 0.8). The use of OTC combination sleep and pain medications was unchanged. Conversely, days in which only Rx medications were used declined in CFS (OR = 0.9), as did the proportion of days on which only single‐ingredient OTC medications were taken (OR = 0.8).

Table 3.

Usage of different product types on usage days, by season

	% of usage days each medication type was used, by season
Medication types	Off‐season (n = 13 121 days)	Cold/flu (n = 38 136 days)	ORa (95% CI)
OTC or Rx medications
Rx only	14.1	12.3	0.9 (0.8–1.0)
OTC only	79.9	81.4	1.1 (1.0–1.2)
Both OTC and Rx	6.1	6.3	1.0 (0.9–1.2)
Classes of OTC medications
OTC single ingredient medications only	38.8	34.8	0.8 (0.8–0.9)
OTC combination medications only	41.3	45.6	1.2 (1.1–1.3)
Both single and combination OTC medications	5.8	7.3	1.3 (1.1–1.5)
Sub‐classes of OTC combination medications b
Pain	13.3	13.0	1.0 (0.9–1.1)
Sleep	7.9	7.8	1.0 (0.9–1.2)
Allergy	12.0	9.3	0.8 (0.7–0.9)
Upper Respiratory	24.8	33.2	1.6 (1.5–1.7)
Upper Respiratory only	17.3	23.9	1.6 (1.4–1.7)
Both Upper Respiratory and other medications	7.5	9.3	1.3 (1.1–1.5)

Entries are percentages of acetaminophen usage days in each season

^aOdds ratios and confidence intervals based on GEE, accounting for nesting of days within persons.

^bPercentages by category of OTC combination medications do not sum to the totals for use of OTC combination medications, because multiple classes may have been used and because some medications may belong to more than one class.

Based on data from March 2011 to March 2016

Bold face ORs and confidence intervals indicate statistically significant seasonal differences

To better understand seasonal increases in the use of upper respiratory medications, we distinguished the sole use of these medications on a given day from use along with other paracetamol medications within the same day. In both CFS and OFFS, on two‐thirds of days when upper respiratory medications were used, they were the only paracetamol medications used. Both patterns of upper respiratory medication use increased in CFS: the odds that a usage day included use of both upper respiratory medications and other paracetamol medications increased by 30% in CFS (Table 3), and the odds that upper respiratory medications were used alone increased even more steeply by 60%.

The observed increases in usage of upper respiratory medications during CFS suggest that usage of these medications could account for seasonal increases in the likelihood of exceeding 4 g. To assess this hypothesis, we re‐computed daily paracetamol intake while excluding intake from upper respiratory medications. As shown in Table 2, this largely eliminated the seasonal increase in exceeding 4 g during CFS (OR = 1.1, not significant, for both persons and days). Further analysis showed that the seasonal increase in exceeding 4 g was due both to days when upper respiratory medications were used exclusively and to days when they were used along with other medications (results not shown).

Discussion

Using detailed diary data covering 5 years and over 14 000 paracetamol users studied over 59 000 usage days, the present analyses showed that 6.3% of paracetamol users exceeded the maximum daily dose on at least one day, with such overdosing occurring on 3.7% of paracetamol usage days. In a previous study with very similar methods, we reported lower prevalence of intake exceeding 4 g: 4.5% of users 11 and 3.1% of usage days 10. The previous study was based on data collected during a single summer month, whereas the current data cover 5 years of data encompassing seasonal variations and a longer time‐frame. Thus, we regard the current estimates as more robust.

Consistent with other studies 20, 21, 22, 23, we focused on exceeding 4 g a day as our primary outcome because this has been designated as the maximum daily dose by the Food and Drug Administration (FDA). This makes the 4‐g limit an apt target for consumer education; also, if users limited their intake to 4 g a day, exposures to still higher doses would necessarily be avoided. However, the dose that can cause liver injury is generally much higher than this and is likely to vary with genetic and other factors 24; repeated instances of excess dosing may also increase the risk. So, there is no implication that the study participants whose reported use exceeded 4 g experienced any adverse medical consequences. We observed that exceeding 8 g in a day was much less common, as was exceeding 4 g on the majority of days (at least 4 out of 7). Indeed, while we documented increased incidence of doses exceeding 4 g on any one day, or on most days, during CFS, the study lacked the power to show similar increases in dosing exceeding 8 g per day, and Weinstein et al. 25 found no seasonal increase in paracetamol‐related liver injury.

It should be noted that there have been several initiatives that may over time reduce excess intake of paracetamol. Starting in 2014, the FDA limited the dose of paracetamol in Rx paracetamol–opiate combinations to 325 mg 26. Further, both tighter prescribing restrictions on opiates 27 and a campaign against over‐prescribing 28 have been fielded, out of concern for increasing overdose deaths, and prescribing is in fact declining 29. Several public‐education campaigns promoting safe use of paracetamol have also been mounted 30, 31. Analyses of trends over time may give insight into whether these interventions have indeed reduced the likelihood of paracetamol dosing exceeding the recommended daily limit of 4 g.

Our analyses also documented seasonal variation in use exceeding the 4‐g limit. The absolute increases in exceeding 4 g during the cold and flu season were modest – just over 1% of persons and days – but represented meaningful relative increases (24% for persons, 37% for usage days and 67% for exceeding 4 g on most days) over the off‐season rates. The increases were almost entirely attributable to seasonal increases in the frequency of use of upper respiratory medications – OTC combination paracetamol products used to treat upper respiratory cold and flu symptoms – which were more often reported during CFS. Both days when only upper respiratory medications were used and days in which upper respiratory medications were used along with other paracetamol medications contributed to the seasonal increase in likelihood of exceeding the daily maximum dose.

In interpreting our reported rates of exceeding the daily 4‐g limit, it is important to note that they are not general‐population rates: as in our previous analyses 10, 11, but unlike analyses in other studies 21, 22, 23, the rates are based on individuals who had used paracetamol in the 30 days preceding the study, and who also used it during the week of diary‐keeping. Thus, the present rates are based on a sample of relatively frequent users of paracetamol, and probably over‐estimate the overall population rates of exceeding 4 or 8 g paracetamol in a day. The screening for individuals who had used paracetamol in the 30 days preceding the study may also have had differential effects during CFS, in that it would tend to exclude those who used paracetamol only very sporadically, such as when they suffer the occasional cold or flu, thus potentially under‐estimating seasonal variation. Conversely, it is possible that sporadic users who take paracetamol medications when they suffer seasonal illnesses might be more prone to use the medications to excess. Note also that the daily rates of exceeding 4 g do not reflect all days, as they are based on days when paracetamol was used, leaving out more than half the total diary days.

The findings suggest the need for public education campaigns to discourage paracetamol dosing that exceeds the recommended maximum dose. Previous findings 10 suggested a need to emphasize avoidance of concomitant use of multiple paracetamol medications.

Our findings suggest that public education campaigns might also include some emphasis on CFS, addressing the use of upper respiratory medications to treat seasonal respiratory illnesses.

A limitation of the study was its use of a sample recruited from internet research panels, which may not be fully representative, particularly underrepresenting persons with low education and limited health literacy 11, who may be more likely to exceed the 4‐g daily limit 32. Response rates were low, which could have introduced nonresponse bias. The data are based on self‐report, which is subject to a number of distortions. The analyses relied on Google searches to identify CFSs. Search behaviour may be influenced by factors besides actual disease incidence that affect consumer concern, but consumer concern may be critically important, since it influences medication‐taking, perhaps even more so than objective measures of illness. Our examination was limited to CFS, and did not encompass other potential seasonal influences, such as allergy seasons, which may introduce geographic as well as seasonal variation. As noted, our analyses were based on individuals who had used paracetamol medications in the month preceding the study, so the sample is likely to over‐represent regular users. The behaviour of more intermittent users may differ; in particular, they may have a lower likelihood of exceeding the daily 4‐g limit. Finally, although exceeding the 4‐g limit is an important behavioural end‐point, its relation to injury is not clear. Use of 4 g per day, even for extended periods, is generally well tolerated 33; it is likely that higher doses over multiple days are necessary to produce liver injury 34.

The study also had several strengths. Definitions of CFSs were validated independently by market sales data and by participants' reports of symptoms. The data were based on detailed diary reports, completed daily by a large national sample, covering 5 years. Participants did not know the study concerned paracetamol and did not need to know which medications contained paracetamol, nor was their attention drawn to seasonal effects, so self‐reports would not be influenced by beliefs about paracetamol or seasonality. Furthermore, with daily completion of the diary, the recall period was short and conducive to accurate reporting. Finally, the data were collected online, which is more likely to elicit truthful reporting 35.

Conclusions

We found that 6.3% of paracetamol users exceeded the maximum recommended daily dose of paracetamol at least once during a week in which they used paracetamol, and 1.4% exceeded this limit on the majority of days; the limit was exceeded on 3.7% of paracetamol usage days. This was somewhat more frequent during CFS. The seasonal increase was mainly attributable to increases in the use of OTC paracetamol medications containing a combination of ingredients designed to treat upper respiratory symptoms, in addition to treating pain and fever. These findings suggest the need for consumer education about the proper use of paracetamol medications, including limiting the dose to that specified in the medication directions, with some special emphasis during CFS on limiting intake of medications treating upper respiratory symptoms.

Competing Interests

The study was ruled exempt by the Boston University Medical Campus Institutional Review Board. Participants consented when completing the surveys. S.S., D.R.B. and Pinney Associates are consultants to Johnson & Johnson Consumer, Inc. and have consulted to other companies that market competing OTC analgesics. J.P.K. has no competing interests to declare. M.K.M. is a consultant to Johnson & Johnson Consumer, Inc. R.B.W. is an employee of Janssen Research and Development, LLC. Janssen Research and Development is part of Johnson & Johnson. D.W.K. received research support from Bayer during the conduct of the study and has been a consultant to UCB.

The authors acknowledge the efforts of the staff at Burke Research, which carried out the data collection. The authors are grateful for valuable guidance from members of a board of advisors for the study: Neeraj Aurora (University of Wisconsin‐Madison), Eric Brass (University of California, Los Angeles), Heidi Jolson (independent consultant) and Leslie Shifrin (independent consultant). The authors also acknowledge the assistance of Mark Sembower, Steve Pype (Pinney Associates), Michael Bairos (Slone Epidemiology Center) with data management and analysis, and Katherine Kelley (Slone Epidemiology Center) with pharmacy‐related issues.

A version of this analysis was presented as a poster at the 32nd International Conference on Pharmacoepidemiology and Therapeutic Risk Management, 25–28 August 2016, Dublin, Ireland. This research was sponsored by Johnson & Johnson Consumer, Inc., which markets Tylenol‐brand paracetamol medicines. Decisions about the content and submission were made solely by the authors.

Contributors

S.S. and D.W.K. conceptualized the study. D.R.B. performed the statistical analyses. All authors oversaw the study and participated in interpretation of the study. S.S. drafted the manuscript, and all authors made substantial contributions to its development.

Supporting information

Appendix S1 Daily symptoms reported, by season

Appendix S2 Percentage of paracetamol dosing occasions used to treat various symptoms

Appendix S3 Comparison of user characteristics previously associated with exceeding 4 g, by season

Shiffman, S. , Battista, D. R. , Kelly, J. P. , Malone, M. K. , Weinstein, R. B. , and Kaufman, D. W. (2018) Prevalence of exceeding maximum daily dose of paracetamol, and seasonal variations in cold‐flu season. Br J Clin Pharmacol, 84: 1250–1257. doi: 10.1111/bcp.13551.