Ambroxol has multiple mechanisms of action; while most of them including secretolytic, anti-inflammatory and anti-oxidant activity are assumed to occur by systemic exposure , Na+ channel blockade in the treatment of sore throat is assumed to occur by a local effect [9, 10]. However, symptom reduction by syrup and pastilles may partly also involve local effects, for instance in the pharynx as part of the cough inhibition. Against this background, we have surveyed pharmacy customers obtaining different formulations of ambroxol as over-the-counter medication to explore specific profiles of patient groups selecting these formulations as well as their corresponding efficacy and tolerability in a real-world setting.
Critique of methods and feasibility
Our data are based on a survey of pharmacy customers purchasing one of four branded ambroxol containing products. The resulting data are not expected to have the same quality as those collected by a physician or other healthcare professional. Moreover, the use of anonymous reponses implies that source data verification was not possible. However, we deem this setting suitable to obtain real-world evidence for over-the-counter medications. Moreover, our previous work using other airway-related over-the-counter medications has demonstrated the external validity of this approach for generating real-world evidence for non-prescription medicines [19, 20].
Our study design did not include a control group, e.g. placebo, for three reasons: Firstly, the efficacy and tolerability of ambroxol has been demonstrated in numerous controlled trials . Therefore, there was no need to re-establish this using a placebo group. Second, such a control group would have interfered with the non-interventional character of our study. Third, the primary intention of our survey was the comparison between ambroxol formulations. Therefore, our data should not be interpreted as proof of efficacy or tolerability but rather as complementary to previously reported controlled studies. It flows from the non-interventional character of our study that we do not have specific data on ingested doses. While we can assume that ambroxol administration was in line with dosing recommendations of the package insert, this limits comparison of formulations based on exposure data. As the primary aim of the study was explorative, no a priori hypotheses existed; accordingly, no hypothesis-testing statistical tests were applied based on recent recommendations .
We have used a patient-adapted version of the BSS to obtain key data (Table 2). The BSS has been validated as a tool for controlled trials [16, 17]. One of the items in the BSS is “rales on auscultation”, which obviously needs to be assessed by a healthcare professional. As this would counter the intention to generate real-world evidence, we have replaced this item by the subjective patient-assessed symptom of “rattles”. Our data show that pharmacy customers can use this adapted version of the BSS as a self-administered tool.
Based on the efficacy and safety of ambroxol in controlled studies in both adults and children [4, 11,12,13], our study included a pediatric syrup which yielded somewhat different results than the adult syrup, particularly for time to onset of symptom relief. Several factors may have contributed to such reported differences. Firstly, the vast majority of the users were children and adolescents, but some users have been adults as indicated by the age of 12.8 ± 12.8 years of users of the pediatric syrup (Table 1). This does not represent inappropriate use because the pharmacologically active content is identical to the adult syrup and the package insert of the pediatric syrup also includes dosing recommendations for adults. Nonetheless, it limits interpretation of these data. Second, users of the pediatric syrup tended to have less severe baseline symptoms than those of the adult syrup. Third, the pathophysiology of cough may differ between children and adults. Forth, and perhaps most importantly, data on children were not reported by the patient but by their parents. Previous studies have shown that parental reporting of children’s cough may be biased, for instance smoking parents under-reported night-time cough in their children . However, parent reporting of effects of cough on the quality of life in children has been validated as reliably according to other studies . Therefore, interpretation of data obtained with the pediatric syrup needs to consider potential bias from inclusion of some adult users, differences in baseline severity as well as reporting bias.
Among pooled users of all four ambroxol formulations, cough with moderate and with much sputum production were the most frequently mentioned lead symptoms, accounting for more than 70% of all patients, whereas only 13% reported dry cough as their lead symptom. Accordingly, patients rated the severity of cough higher than that of any other item of the BSS (Table 2). This is not surprising because 45% of participants reported to experience an average of four or more coughs per hour during daytime (Fig. 1). Similarly, 36 and 35% reported “full” or “mostly” impairment of falling asleep (Fig. 3). Awakening four or more times during the night due to coughing was reported in less than 25% of subjects (Fig. 2). While this indicates that coughing impairs falling asleep more frequently than staying asleep, it should be noted that patients experience nighttime cough as a much greater burden than daytime cough . In contrast, users of ambroxol reported full impairment of their ability to concentrate or to execute daily tasks in only 18 and 15% of cases, respectively.
Among formulations for intended use in adults, there was a consistent trend that users of soft pastilles had the least severe symptoms at baseline, be it for the adapted BSS or any of the symptoms shown in Figs 1, 2, 3, 4, 5 and 6. This may reflect that pastilles generally are considered a rather mild form of treatment, perhaps even seen as acting purely locally, whereas capsules and adult syrup are perceived to act systemically. Such perception may have caused a selection bias among those with less severe symptoms.
The present data as reported by pharmacy customers are in line with those from controlled clinical studies [4, 11,12,13], confirming the efficacy and tolerability of ambroxol in a real-world setting. Moreover, they demonstrate that the patient-adapted version of the BSS is suitable as a self-assessment tool. However, in the absence of reports on effects of ambroxol on BSS in controlled trials, this does not substitute for a formal validation of this version.
The primary aim of the present study had been a comparison of four ambroxol formulations. Focusing on the BSS as efficacy parameter and considering differences in baseline severity among user groups, efficacy was comparable between the formulations. If anything, the pediatric syrup was reported to be slightly more effective and the adult syrup slightly less effective. A similar pattern was also observed for the questions on impairment by the respiratory symptoms. Several factors may have contributed to minor differences in reported efficacy. These include local effects of syrups and pastilles related to local anesthetic action [9, 10], antitussive effects of syrups irrespective of active pharmacological ingredients [15, 24] as well as selection and reporting biases; factors specific for users of the pediatric syrup have been discussed above.
About 90% of participants reported the onset of effect to require no more than 60 min, showing a fast onset of all formulations. This is earlier than the systemic pharmacokinetic tmax of the ambroxol formulations, which is 1–2.5 h for all immediate release formulations  and 6.5 h for ER capsules . This difference may in part be explained by local effects of the syrups and pastilles, but also points to a possible placebo component in reported time to onset of symptom relief.
Time to onset of symptom relief is the parameter differing most notably between formulations. Thus, compared to the adult syrup, users of pediatric syrup and pastilles reported a somewhat faster and those of the ER capsules a somewhat slower onset of symptom relief. A slower onset of action of the ER capsules is compatible with their later systemic pharmacokinetic tmax (6.5 vs. 1–2.5; ). A faster onset of action after ingestion of first dose with pastilles may be explained by a longer local contact time in the pharynx, allowing for a greater contribution of local anesthetic effects [9, 10]. Factors possibly involved in specific differential findings in users of the pediatric syrup have been discussed above.
Across all formulations, reported efficacy of treatment increased to some degree with duration of use, i.e. efficacy was rated as very good by 31.0% of subjects using ambroxol for 1 day only, with a step-wise increase to 38.0% in 5-day users. Based on a terminal elimination half-life of 7–11 h [25,26,27], pharmacokinetic steady-state is not expected to occur earlier than late on the second day of treatment, making earlier full efficacy unlikely. On the other hand, subjects stopping treatment after the first day (other than for adverse events, a situation not reported in this study) are likely to represent a biased sample of patients with very good symptom improvement. As symptoms of acute bronchitis may spontaneously resolve within few days, it is difficult to assess the relative roles of bias vs. reaching pharmacokinetic steady state in the absence of a placebo control group.
Tolerability assessment in a non-controlled study is difficult due to lack of a control group. This particularly applies if any worsening of symptoms or condition-associated impairments is counted as AE, as was done in our study. Such worsening accounted for most of the reported AEs. This is reflected by 97.3% of all participants rating global tolerability as very good or good and the overall improvements in symptom and associated impairment scores at the group level. Therefore, the tolerability data from the present study are not providing new signals or concerns relative to the established tolerability and safety of ambroxol.