Abstract
Maintaining medication adherence remains a major clinical challenge, as higher daily dosing frequencies are often associated with decreased adherence. Although once-daily regimens are generally preferred, specific patient preferences regarding dosing frequency are unclear, despite implications for improving patient satisfaction and optimizing pharmacotherapy. Here, we evaluated the frequency at which patients begin to perceive dosing as excessive. A web-based questionnaire survey using the personal health record infrastructure of electronic medication notebooks was administered between July 20 and 26, 2023. Eligible participants were aged ≥20 years and had received oral tablets or capsules within 90 d of the survey. The questionnaire consisted of 6 items, including whether participants felt that the daily medication frequency was excessive. A multivariate logistic regression analysis was performed to identify the frequency beyond which patients perceive dosing as excessive, adjusting for confounders, such as medication formulation, number of medications, and patient characteristics. Of 1478 respondents, 1236 were included in the analysis. In total, 28.9% of participants reported that their medication frequency felt excessive. In the multivariate logistic regression analysis, twice-daily or more frequent dosing was significantly associated with the perception of excessiveness, using once-daily dosing as the reference. Twice-daily dosing is the frequency beyond which patients are significantly more likely to perceive medication as excessive. Our findings emphasize the importance of simplifying dosing regimens.
INTRODUCTION
Medication adherence is a major clinical issue. Although various factors influence adherence, it is self-evident that patients experience greater stress as the number of doses per day increases. Many studies have shown that medication adherence rates for twice-, three-times-, and four-times-daily dosing regimens are significantly lower than those for once-daily regimens.1–3) Furthermore, Mohammed et al. demonstrated that medication-related burden affects patients’ well-being and their beliefs and behaviors toward medicines.4) However, the specific daily frequency at which patients begin to perceive dosing as excessive is unknown.
Questionnaire-based surveys provide a methodological approach for elucidating the frequency beyond which dosing is perceived as excessive. However, conventional questionnaire survey methods have limitations, including difficulty in accurately obtaining detailed information, such as prescription data. Given that various factors, including the size, formulation, and complexity of medications, in addition to dosing frequency itself, contribute to the perceived burden of medication use, comprehensively accounting for these variables is essential. In other words, conventional methods limit the depth of analysis.
Herein, we utilized the “electronic medication notebook” as a survey platform. It is currently a popular digital version of the traditional medication notebook that uses technologies such as smartphones.5) It functions as a personal health record (PHR), allowing centralized management of dispensed medication details. Furthermore, questionnaires can be directly distributed to patients via this platform, enabling the simultaneous collection of PHR data, including dispensing histories, alongside survey responses. We previously investigated patient preferences regarding oral medications based on electronic medication notebooks, identifying various factors, such as tablet size, that contribute to swallowing difficulties and challenges in removing tablets or capsules from press-through-package sheets.6–8)
In this study, leveraging this platform, we determined the daily medication frequency at which patients begin to perceive dosing as excessive compared with a once-daily regimen.
MATERIALS AND METHODS
Questionnaire Survey
The questionnaire was administered to users of the electronic medication notebook (harmo®; harmo Co., Ltd., Tokyo, Japan) from July 20 to 26, 2023. The inclusion criteria were: (1) aged ≥20 years and (2) received tablets or capsules within 90 d from a week prior to the survey distribution date (July 12, 2023). We sent notifications about the questionnaire through the harmo smartphone app, including a link to the survey details. Users were asked to provide consent on the platform and to respond to the questionnaire about medications they were currently taking.6–8) The questionnaire contained the following items: Q1, do participants feel the daily medication frequency is excessive (yes/somewhat yes/somewhat no/no); Q2, do participants feel that the amount taken at one time is too much (where “amount taken” refers to the number of tablets or capsules or the amount of liquid or powder medicine; yes/somewhat yes/somewhat no/no); Q3, what is the timing of the dose (if the response to Q2 was either yes or somewhat yes; after breakfast/after lunch/after dinner/others; multiple answers were allowed); Q4, are all prescribed medications being taken as instructed (except when forgotten) (I can take it/I adjust the dose according to the clinician’s advice/I adjust the dose using my judgment); Q5, what is the name of the drug and how is it adjusted (if the response to Q4 was I adjust the dose according to the clinician’s advice or I adjust the dose by my own judgment); and Q6, do participants feel stressed or dissatisfied with the use of the current prescription medication based on the previous questions (yes/somewhat yes/somewhat no/no).
Data Collection
Demographic data, including sex, age (grouped by decade), and medication dispensing history, were collected anonymously. Oral medications overlapping 1 week before July 12 were defined as “current medications.” Because determining prescription periods for as-needed medications and topical formulations was difficult, they were excluded from the analysis. Participants who did not have any current medications were excluded. Based on the medication dispensing history, we investigated the number of doses per day, timing of medication administration, number of medications taken per day, formulation type (powdered or liquid), and whether medications were considered large. Additionally, we collected data on the use of specific drug classes: antidepressants [Anatomical Therapeutic Chemical Classification (ATC) code: N06A], anti-Parkinson’s disease drugs (ATC code: N04), anticancer drugs (ATC code: L01), and antidiabetic drugs (ATC code: A10). The number of doses was defined as the total number of administration time points in a day. For example, a patient prescribed 2 tablets after breakfast and 2 tablets after dinner was considered to have 2 daily doses. The number of medications was defined as the total number of tablets and units taken at each administration time point. For example, if a patient was prescribed 2 tablets of drug A, 1 tablet of drug B, and 1 unit of liquid drug C after breakfast, the total number of medications would be 4 (2 tablets + 1 tablet + 1 unit). “Large medications” were defined based on our previous study using the index of “long diameter + short diameter + thickness” at a 21.5-mm threshold, which serves as the limit for perceived swallowing difficulty.6)
Endpoint
The primary endpoint was the daily medication frequency at which participants begin to perceive dosing as excessive, with reference to once-daily dosing.
Statistical Analysis
To identify the medication frequency at which patients perceive dosing as excessive, a multivariate logistic regression analysis was performed. The outcome variable was whether participants felt that the daily medication frequency was excessive. The explanatory variables included the daily medication frequency and several covariates to adjust for confounding: questionnaire results (Q2, Q4, and Q6), demographic characteristics (sex and age), medications (number of doses per day, with powder medications, with liquid medications, with large medications, antidepressant use, anti-Parkinson’s disease drug use, anticancer drug use, and antidiabetic drug use), number of medications, and timing of medication use. Although a factor analysis with a similar focus has not been conducted, we referred to covariates used in previous studies of factors affecting medication adherence.9–11) The variance inflation factor (VIF) was calculated to assess multicollinearity among explanatory variables. The selection of covariates was guided by VIF values (i.e., all included variables had VIFs of less than 5). In this analysis, responses of “yes” and “somewhat yes” were both classified as “yes.” Statistical tests were not performed for comparisons between 2 groups, such as patient background characteristics, in accordance with the American Statistical Association’s Statement.12)
All statistical analyses were conducted using R version 4.5.0 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was set at p < 0.05.
Ethical Considerations
This study involved a web-based self-administered questionnaire survey, and all data were anonymized and de-identified. Only individuals who provided consent to participate in the survey and shared the PHR managed by harmo Co., Ltd. were included. The study protocol was approved by the Ethics Committee of the Faculty of Pharmacy, Keio University (230420-1). No incentives were provided to respondents.
RESULTS
Questionnaire Results and Patient Characteristics
Among 50080 participants who received the questionnaire invitation, 5135 opened the survey request notification, 1793 accessed the questionnaire page, and 1478 responded to the questionnaire. A total of 242 respondents without current medications were excluded, resulting in the inclusion of 1236 participants (Fig. 1). Regarding Q1 (whether the daily medication frequency is excessive), 357 participants (28.9%) answered “yes (132 participants)” or “somewhat yes (225 participants).” On the other hand, 879 participants (71.1%) answered “no (405 participants)” or “somewhat no (474 participants).” Detailed results for participants who “do or do not feel that the number of medications per day is excessive” are shown in Table 1 (excluding Q5). The proportion of respondents who answered “yes” or “somewhat yes” to Q2 (do participants feel that the amount taken at one time is too much) and Q6 (do participants feel stressed or dissatisfied with the use of the current prescription medication) was higher in respondents who indicated that the medication frequency is excessive (excessive group) than in those who did not. Regarding Q5 (the name of the drug and how it is adjusted), over half of the respondents in both groups reported self-adjusting medications, such as analgesics, laxatives, probiotics, and sleeping agents. Sex distribution was similar in the excessive and not excessive groups, and most participants were in their 50s or 60s (Table 2). The daily medication frequency was also higher in the excessive group.
Table 1. Detailed Questionnaire Results between Participants Who Do or Do Not Feel That the Number of Medications per Day Is Excessive
| Questionnaire results, n (%) |
Excessive |
Not excessive |
| Q2: Do participants feel that the amount taken at one time is too much, n (%) |
N = 357 |
N = 879 |
| Yes |
146 (40.9) |
62 (7.1) |
| Somewhat yes |
128 (35.9) |
159 (18.1) |
| Somewhat no |
63 (17.6) |
325 (37.0) |
| No |
20 (5.6) |
333 (37.9) |
| Q3: What is the timing of the dose, n (%) |
N = 274 |
N = 221 |
| After breakfast |
232 (84.7) |
183 (82.8) |
| After lunch |
51 (18.6) |
13 (5.9) |
| After dinner |
168 (61.3) |
90 (40.7) |
| Others |
44 (16.1) |
26 (11.8) |
| Q4: Are all prescribed medications being taken as instructed, n (%) |
N = 357 |
N = 879 |
| I can take it |
319 (89.4) |
814 (92.6) |
| I adjust the dose according to the clinician’s advice |
18 (5.0) |
35 (4.0) |
| I adjust the dose by my own judgment |
20 (5.6) |
30 (3.4) |
| Q6: Do participants feel stressed or dissatisfied with the use of the current prescription medication, n (%) |
N = 357 |
N = 879 |
| Yes |
33 (9.2) |
23 (2.6) |
| Somewhat yes |
132 (37.0) |
86 (9.8) |
| Somewhat no |
147 (41.2) |
350 (39.8) |
| No |
45 (12.6) |
420 (47.8) |
Excessive or not excessive: Participants who feel that the number of medications per day is excessive or not. Q2: “Amount taken” refers to the number of tablets or capsules or the amount of powder or liquid medicine (yes/somewhat yes/somewhat no/no); Q3: if either yes or somewhat yes in Q2 (after breakfast/after lunch/after dinner/others; multiple answers were allowed); Q4: except for cases where medicine was forgotten.
Table 2. Detailed Comparison of Participants Who Feel That the Medication Frequency Is Excessive or Not
|
Excessive (n = 357) |
Not excessive (n = 879) |
| Sex (female), n (%) |
139 (38.9) |
364 (41.4) |
| Age (years), n (%) |
|
|
| 20s |
9 (2.5) |
21 (2.4) |
| 30s |
27 (7.6) |
61 (6.9) |
| 40s |
56 (15.7) |
125 (14.2) |
| 50s |
103 (28.9) |
220 (25.0) |
| 60s |
95 (26.6) |
282 (32.1) |
| 70s |
52 (14.6) |
145 (16.5) |
| 80s |
14 (3.9) |
23 (2.6) |
| 90s |
1 (0.3) |
2 (0.2) |
| Number of doses per day (median [IQR]) |
3.0 [2.0–4.0] |
2.0 [1.0–3.0] |
| Number of doses per day, n (%) |
|
|
| One |
28 (7.8) |
283 (32.2) |
| Two |
76 (21.3) |
244 (27.8) |
| Three |
98 (27.5) |
158 (18.0) |
| Four |
78 (21.8) |
115 (13.1) |
| Five |
30 (8.4) |
27 (3.1) |
| Six |
47 (13.2) |
52 (5.9) |
| With powder medications, n (%) |
96 (26.9) |
129 (14.7) |
| With liquid medications, n (%) |
15 (4.2) |
14 (1.6) |
| With large medications, n (%) |
240 (67.2) |
431 (49.0) |
| Number of medications per day (median [IQR]) |
9.0 [5.0–14.0] |
5.0 [2.0–8.0] |
| Antidepressants use, n (%) |
150 (42.0) |
272 (30.9) |
| Anti-Parkinson’s disease drug use, n (%) |
13 (3.6) |
14 (1.6) |
| Anticancer drug use, n (%) |
3 (0.8) |
1 (0.1) |
| Antidiabetic drug use, n (%) |
92 (25.8) |
164 (18.7) |
| Timing of medication use, n (%) |
|
|
| At waking |
15 (4.2) |
11 (1.3) |
| Before breakfast |
74 (20.7) |
105 (11.9) |
| After breakfast |
340 (95.2) |
763 (86.8) |
| Between breakfast and lunch |
10 (2.8) |
9 (1.0) |
| Before lunch |
54 (15.1) |
58 (6.6) |
| After lunch |
187 (52.4) |
239 (27.2) |
| Between lunch and dinner |
7 (2.0) |
8 (0.9) |
| Before dinner |
72 (20.2) |
96 (10.9) |
| After dinner |
307 (86.0) |
566 (64.4) |
| Between dinner and bedtime |
9 (2.5) |
10 (1.1) |
| Before bedtime |
179 (50.1) |
307 (34.9) |
| Others |
0 (0.0) |
1 (0.1) |
Excessive or not excessive: Participants who feel that the medication frequency is excessive or not. Large medications were defined as those with a sum of long diameter, short diameter, and thickness of 21.5 mm or more. IQR: interquartile range.
The VIF for the “timing of medication use” exceeded 5; therefore, this variable was excluded from the logistic regression analysis. The VIF values for all other variables were below 5, indicating no multicollinearity (Supplementary Table 1).
In a multivariate logistic regression analysis (Table 3), twice-daily or more frequent dosing was significantly associated with the perception of excessiveness. Additionally, responses to Q2 (do participants feel that the amount taken at one time is too much) and Q6 (do participants feel stressed or dissatisfied with the use of the current prescription medication) were both significantly associated with the perception of excessiveness.
Table 3. Multivariate Logistic Regression Analysis to Identify the Daily Medication Frequency at Which Participants Begin to Perceive Dosing as Excessive
|
Odds ratio (95% CI) |
p-Value |
| Q2: Yes or somewhat yes |
6.63 (4.72–9.39) |
<0.05 |
| Q4: I can take it |
0.94 (0.55–1.63) |
0.83 |
| Q6: Yes or somewhat yes |
3.77 (2.69–5.31) |
<0.05 |
| Sex (female) |
0.78 (0.57–1.08) |
0.131 |
| Age (years) |
|
|
| 20s |
0.65 (0.23–1.75) |
0.405 |
| 30s |
1.01 (0.51–1.97) |
0.984 |
| 40s |
1.00 (reference) |
|
| 50s |
1.29 (0.80–2.11) |
0.303 |
| 60s |
0.66 (0.40–1.08) |
0.094 |
| 70s |
0.68 (0.39–1.20) |
0.185 |
| 80s |
0.98 (0.38–2.45) |
0.959 |
| 90s |
1.08 (0.05–13.56) |
0.955 |
| Number of doses per day |
|
|
| One |
1.00 (reference) |
|
| Two |
2.35 (1.39–4.05) |
<0.05 |
| Three |
4.23 (2.43–7.53) |
<0.05 |
| Four |
4.16 (2.16–8.14) |
<0.05 |
| Five |
4.40 (1.90–10.29) |
<0.05 |
| Six |
4.61 (1.93–11.11) |
<0.05 |
| With powder medications |
0.91 (0.58–1.43) |
0.685 |
| With liquid medications |
1.03 (0.42–2.51) |
0.956 |
| With large medications |
0.77 (0.54–1.11) |
0.162 |
| Antidepressant use |
0.90 (0.63–1.27) |
0.537 |
| Anti-Parkinson’s disease drug use |
1.13 (0.44–2.90) |
0.792 |
| Anticancer drug use |
7.85 (0.63–220.75) |
0.150 |
| Antidiabetic drug use |
0.90 (0.61–1.32) |
0.598 |
| Number of medications per day |
1.00 (0.97–1.03)* |
0.945 |
Q2: Do participants feel that the amount taken at one time is too much (“amount taken” refers to the number of tablets or capsules, or the amount of powder or liquid medicine; yes/somewhat yes/somewhat no/no); Q4: are all prescribed medications being taken as instructed (except when forgotten) (I can take it/I adjust the dose according to the clinician’s advice/I adjust the dose by my own judgment); and Q6: do participants feel stressed or dissatisfied with the use of the current prescription medication (yes/somewhat yes/somewhat no/no). Large medications were defined as those with a sum of long diameter, short diameter, and thickness of 21.5 mm or more. * Odds ratio indicates odds per single unit increase. CI: confidence interval.
DISCUSSION
This study revealed that patients are more likely to perceive medication frequency as excessive when the regimen involves twice-daily dosing or more. Furthermore, the transition from twice- to three-times-daily dosing may contribute to a further increase in perceived excessiveness. While previous research has demonstrated a decline in adherence with increased dosing frequency,1–3) our findings provide evidence of a frequency level at which patients are more likely to experience treatment-related stress and decreased satisfaction. Although pharmacotherapy must be designed by considering multiple clinical factors, including efficacy, safety, disease characteristics, and pharmacokinetics,4) our findings provide insights into reducing medication-related burden, which is strongly associated with patients’ physical and psychological well-being.
A notable strength of this study is the use of the PHR infrastructure via electronic medication notebooks. This platform enabled the accurate evaluation of the status of medications by linking questionnaire surveys to detailed prescription histories.6–8) Unlike conventional surveys, precise estimation of dosing frequency and determination of when patients perceive the regimen as excessive was possible. To control potential confounding, we adjusted for multiple patient- and medication-related factors, including number of medications per day, formulation types, and medication size. We also accounted for responses on the perceived amount of medication, compliance, and general dissatisfaction with medications. The association between a higher dosing frequency and perceived excessiveness remained significant after controlling for these covariates, indicating the robustness of this association.
In a logistic regression analysis, answers of “yes” or “somewhat yes” to Q2 (do participants feel that the amount taken at one time is too much) and Q6 (do participants feel stressed or dissatisfied with the use of the current prescription medication) were significantly associated with the perception of medication frequency as excessive. Regarding Q2, patients who take a large number of medications at one time may be more likely to feel that the frequency is excessive. As for Q6, it remains unclear whether a higher dosing frequency causes decreased treatment satisfaction, or whether lower satisfaction with pharmacotherapy leads to a more perceived burdensome regimen (i.e., whether this represents a causal relationship or a mere correlation). Nonetheless, the finding that these 2 factors are strongly associated with patients’ overall perception of medication burden appears clinically plausible, and their inclusion as covariates in the multivariate analysis was confirmed to be crucial. Although no statistical significance was observed, anticancer drug use showed a high odds ratio, likely due to only four participants using anticancer drugs. Thus, the impact of anticancer drug use warrants further investigation.
This study has several limitations. First, the survey was limited to electronic medication notebook users, potentially introducing selection bias toward digitally literate individuals. Additionally, although Q4 addressed self-reported medication compliance, the proportion of participants who adjusted based on their own judgment (3.4%) was relatively low.9–11) This indicates a potential selection bias toward a population with high medication adherence. Furthermore, most participants were in their 50s and 60s and likely working, which may have influenced the results. Moreover, given the nature of the questionnaire items, it is possible that individuals who were particularly dissatisfied or concerned about their medication use were more likely to respond. Therefore, caution is needed when generalizing these findings. As-needed and topical medications were excluded owing to difficulties in estimating the prescription period, which may have influenced overall medication burden. Additionally, the actual use of medications could not be determined owing to the use of dispensing data. Finally, there is a possibility of unadjusted confounding, including previous medication history. For example, a prior history of different dosing frequencies may have influenced patients’ perceived burden at the time of the questionnaire.
Nevertheless, our study addressed key challenges inherent in traditional survey-based methods and generated clinically relevant insights. The findings may assist healthcare professionals in designing simpler and more acceptable medication regimens aligned with patient perceptions and in promoting sustained treatment adherence.
In conclusion, although once-daily dosing is widely accepted as optimal, transitioning to a twice-daily regimen markedly increases the likelihood of patients perceiving the dosing frequency as excessive. This finding offers a concrete reference point for the design of patient-centered pharmacotherapy.
Acknowledgments
We would like to thank all participants who responded to our survey.
DECLARATIONS
Conflict of Interest
TS and TI are co-CEOs of harmo Co., Ltd., which operates harmo. YI, MT, RK, NY, and MS are employees of harmo Co., Ltd. The remaining authors declare no conflict of interest.
Supplementary Materials
This article contains supplementary materials.
REFERENCES
- 1) Coleman CI, Limone B, Sobieraj DM, Lee S, Roberts MS, Kaur R, Alam T. Dosing frequency and medication adherence in chronic disease. J. Manag. Care Pharm., 18, 527–539 (2012).
- 2) Inoue Y, Oka S, Yokoyama S, Hasegawa K, Mahlich J, Schaede U, Habuka N, Murata Y. Medication adherence of people living with HIV in Japan—a cross-sectional study. Healthcare, 11, 451 (2023).
- 3) Ishida N, Tokumoto Y, Suga Y, Noguchi-Shinohara M, Abe C, Yuki-Nozaki S, Mori A, Horimoto M, Hayashi K, Iwasa K, Yokogawa M, Ishimiya M, Nakamura H, Komai K, Matsushita R, Ishizaki J, Yamada M. Factors associated with self-reported medication adherence in Japanese community-dwelling elderly individuals: the Nakajima study. Yakugaku Zasshi, 141, 751–759 (2021).
- 4) Mohammed MA, Moles RJ, Chen TF. Medication-related burden and patients’ lived experience with medicine: a systematic review and metasynthesis of qualitative studies. BMJ Open, 6, e010035 (2016).
- 5) Orii T. Current status and issues of electronic information provision for risk minimization: coordination between electronic medicine notebook and patient drug information. Yakugaku Zasshi, 138, 325–329 (2018).
- 6) Asano M, Imai S, Shimizu Y, Kizaki H, Ito Y, Tsuchiya M, Kuriyama R, Yoshida N, Shimada M, Sando T, Ishijima T, Hori S. Factor analysis of patients who find tablets or capsules difficult to swallow due to their large size: using the personal health record infrastructure of electronic medication notebooks. J. Med. Internet Res., 26, e54645 (2024).
- 7) Tsuchiya M, Imai S, Asano M, Shimizu Y, Kizaki H, Ito Y, Tsuchiya M, Kuriyama R, Yoshida N, Shimada M, Sando T, Ishijima T, Hori S. Exploring the factors associated with difficulties in extracting tablets or capsules from press-through-package sheets. Br. J. Clin. Pharmacol., 91, 479–484 (2025).
- 8) Shimizu Y, Tsuchiya M, Asano M, Imai S, Kizaki H, Ito Y, Tsuchiya M, Kuriyama R, Yoshida N, Shimada M, Sando T, Ishijima T, Hori S. Analysis of factors affecting difficulty in handling oral medicine using electronic medication notebook-based personal health records. Sci. Rep., 15, 26867 (2025).
- 9) Le Bozec A, Korb-Savoldelli V, Boiteau C, Dechartres A, Al Kahf S, Sitbon O, Montani D, Jaïs X, Guignabert C, Humbert M, Savale L, Chaumais MC. Medication adherence, related factors and outcomes among patients with pulmonary arterial hypertension or chronic thromboembolic pulmonary hypertension: a systematic review. Eur. Respir. Rev., 33, 240006 (2024).
- 10) King K, Czuber-Dochan W, Chalder T, Norton C. Medication non-adherence in inflammatory bowel disease: a systematic review identifying risk factors and opportunities for intervention. Pharmacy, 13, 21 (2025).
- 11) Yap AF, Thirumoorthy T, Kwan YH. Systematic review of the barriers affecting medication adherence in older adults. Geriatr. Gerontol. Int., 16, 1093–1101 (2016).
- 12) Wasserstein RL, Lazar NA. The ASA’s statement on p-values: context, process, and purpose. Am. Stat., 70, 129–133 (2016).
