Scolaris Content Display Scolaris Content Display

Fármacos de venta libre (FVL) para la tos aguda en niños y adultos en entornos comunitarios

Collapse all Expand all

Resumen

Antecedentes

La tos aguda debido a la infección de las vías respiratorias altas (IVRA) es un síntoma frecuente. Los fármacos sin prescripción de venta libre se recomiendan con frecuencia como tratamiento de primera línea, pero hay poca evidencia sobre si estos fármacos son efectivos.

Objetivos

Evaluar los efectos de los preparados orales de venta libre para la tos aguda en niños y adultos en entornos comunitarios.

Métodos de búsqueda

Se hicieron búsquedas en CENTRAL (2014, número 1), MEDLINE (enero 1966 hasta marzo, semana 3, 2014), EMBASE (enero 1974 hasta marzo 2014), CINAHL (enero 2010 hasta marzo 2014), LILACS (enero 2010 hasta marzo 2014), Web of Science (enero 2010 hasta marzo 2014) y en el UK Department of Health National Research Register (marzo 2010).

Criterios de selección

Ensayos controlados aleatorizados (ECA) que comparan las preparaciones orales para la tos de venta libre con placebo en niños y adultos que sufren de tos aguda en ámbitos comunitarios. Se consideraron todos los resultados de la tos; los resultados secundarios de interés fueron los efectos adversos.

Obtención y análisis de los datos

Dos autores de la revisión examinaron de forma independiente las citas potencialmente pertinentes, extrajeron los datos y evaluaron la calidad de los estudios. Cuando fue apropiado, se realizaron análisis cuantitativos.

Resultados principales

Debido a los números pequeños de ensayos en cada categoría, los datos cuantitativos limitados disponibles y las diferencias pronunciadas entre los ensayos en cuanto a los participantes, las intervenciones y la medición de resultados, se consideró que el agrupamiento de los resultados era inapropiado.

Se incluyeron 29 ensayos (19 en adultos, 10 en niños) con 4835 pacientes (3799 adultos y 1036 niños). Todos los estudios fueron ECA controlados con placebo. Sin embargo, la evaluación del riesgo de sesgo de los estudios incluidos se vio limitada por la deficiente presentación de informes, en particular en el caso de los estudios anteriores.

En los estudios con adultos, seis ensayos compararon los antitusivos con el placebo y tuvieron resultados variables. Tres ensayos compararon el expectorante guaifenesina con el placebo; uno indicó un beneficio significativo, mientras que los otros dos no lo hicieron. Un ensayo halló que un mucolítico redujo la frecuencia de la tos y las puntuaciones de los síntomas. Dos estudios examinaron las combinaciones entre los antihistamínicos y los descongestivos y hallaron resultados incompatibles. Cuatro estudios compararon otras combinaciones de los fármacos con placebo e indicaron algún beneficio en la reducción de los síntomas de la tos. Tres ensayos hallaron que los antihistamínicos no fueron más efectivos que el placebo para aliviar los síntomas de la tos.

En los estudios con niños, los antitusivos (datos de tres estudios), los antihistamínicos (datos de tres estudios), los antihistamínicos descongestionantes (dos estudios) y las combinaciones de antitusivos y broncodilatadores (un estudio) no fueron más efectivos que el placebo. Ningún estudio que utilizó expectorantes cumplió con los criterios de inclusión. Los resultados de un ensayo favorecieron el tratamiento activo con mucolíticos sobre el placebo. Un ensayo probó dos jarabes pediátricos para la tos y ambas preparaciones demostraron una “respuesta satisfactoria” en un 46% y un 56% de niños comparados con un 21% de niños en el grupo placebo. Un nuevo ensayo indicó que tres tipos de miel eran más efectivos que el placebo en un período de tres días.

Veintiún estudios informaron sobre los efectos adversos. Hubo una amplia gama de estudios, con un mayor número de efectos adversos en los participantes que tomaban preparados que contenían antihistamínicos y dextrometorfano.

Conclusiones de los autores

Los resultados de esta revisión deben interpretarse con cautela porque el número de estudios en cada categoría de las preparaciones para la tos fue pequeño. La disponibilidad, la dosificación y la duración del uso de los fármacos de venta libre contra la tos varían considerablemente en los distintos países. Muchos estudios fueron mal informados, lo que dificultó la evaluación del riesgo de sesgo, y los estudios también fueron muy diferentes entre sí, lo que dificultó la evaluación de la eficacia general. No existe evidencia convincente a favor o en contra de la efectividad de los fármacos de venta libre en la tos aguda. Esto debe tenerse en cuenta al considerar la posibilidad de prescribir antihistamínicos y agentes antitusivos centralmente activos en los niños; fármacos que se sabe que tienen el potencial de causar daños graves.

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Resumen en términos sencillos

Fármacos de venta libre (FVL) para la tos aguda en niños y adultos en entornos comunitarios

Pregunta de la revisión

Se realizó una revisión para determinar la efectividad de los fármacos de venta libre para la tos en la reducción de la tos en niños y adultos en entornos comunitarios. Se encontraron 29 ensayos con 4835 pacientes.

Antecedentes

La tos aguda es un síntoma común y problemático en los niños y adultos que sufren de una infección aguda de las vías respiratorias altas (IAVRA). Muchos pacientes se automedican las preparaciones para la tos de venta libre y los profesionales de la salud a menudo recomiendan su uso para el tratamiento inicial de la tos. Hay una variación sustancial entre los países en cuanto a la disponibilidad y las guías de uso de muchos de estos preparados.

Características de los estudios

Se identificó una amplia gama de estudios de diferentes tipos de preparados utilizados en diferentes dosis tanto en adultos como en niños.

Resultados clave

La evidencia está actualizada hasta marzo de 2014. No se encontró ninguna buena evidencia a favor o en contra de la efectividad de los fármacos de venta libre en la tos aguda. Diecinueve estudios informaron sobre los efectos adversos de estos fármacos y describieron efectos secundarios poco frecuentes, principalmente menores, como náuseas, vómitos, dolor de cabeza y somnolencia.

Calidad de la evidencia

Los resultados de esta revisión deben interpretarse con cautela porque el número de estudios en cada categoría de las preparaciones para la tos fue pequeño. Muchos estudios fueron mal informados, lo que dificulta la evaluación del riesgo de sesgo. Si bien todos los estudios eran ensayos controlados aleatorizados controlados con placebo, sólo una minoría informó sobre sus métodos de asignación y aleatorización y no se informó sobre el cegamiento de los evaluadores de resultados ni sobre si se habían validado las medidas de resultado de la tos. Además, los estudios apoyados por las empresas farmacéuticas u otros proveedores tenían más probabilidades de arrojar resultados positivos. Los estudios fueron muy diferentes entre sí en cuanto a los tipos de tratamiento, la duración del mismo y los resultados medidos, lo que dificulta la evaluación de la efectividad general de los fármacos de venta libre para la tos.

Authors' conclusions

Implications for practice

There is no good evidence for or against the effectiveness of over‐the‐counter (OTC) cough medicines, and from the studies included in this review it remains unclear whether these medications are helpful for the treatment of acute cough. Although a number of randomised controlled trials (RCTs) have compared OTC cough preparations with placebo, the number of trials in each group was small. This review suggests that most preparations appear to be safe, based on those studies reporting side effects, which only described a low incidence of mainly minor adverse effects. However, more serious concerns about the safety of OTC cough medicines have arisen since this review was last updated, particularly in young children and, in general, larger numbers of patients are required in order to identify serious, though less common adverse effects (Smith 2008a). This systematic review confirms the lack of evidence for or against an effect of OTC cough preparations despite using an extensive search strategy. This lack of evidence of effectiveness also has implications for the regulatory bodies and brings into question how these products can continue to be promoted using language that implies that their effectiveness is not in doubt.

The results of this review have to be interpreted with caution because study designs, populations, interventions and outcomes varied markedly between studies, limiting the generalisability of the results. All results were based on a small number of studies. It is also questionable as to whether all of the positive results obtained with unclear outcome measures are clinically relevant.

Implications for research

Further high‐quality RCTs of OTC cough preparations are needed as the results of this review are based on a small number of often underpowered studies. More evidence about the effectiveness of OTC cough preparations would be helpful, particularly in relation to the use of honey in adults and children, as identification of effective self care treatments may help reduce the burden of days lost at work due to acute cough as well as the number of consultations in primary care. Research should also include individuals who self medicate with OTC cough preparations, as there is likely to be a variation between countries in the proportion of individuals using these medications, with or without professional advice, particularly given the international variation in what products are available OTC or on a prescription basis. There is also a need to identify ineffective preparations in order to lower costs for consumers and healthcare providers. Studies will need to be rigorously designed and should use clinically relevant outcome measures, including cough frequency, severity and duration. It is important that future RCTs use OTC drugs in doses recommended by the manufacturers for an appropriate length of time, as drugs tested in a single and possibly too low a dose are likely to be ineffective. Trials should also report details on effect sizes and provide data on adherence and adverse effects. This review also highlights a need for an outcome measure for acute cough that is clinically relevant, valid, reliable and easy to use in RCTs.

Background

Description of the condition

Acute cough due to upper respiratory tract infection (URTI) is one of the most common symptoms worldwide. A large number of people self prescribe non‐prescription over‐the‐counter (OTC) cough medicines for themselves or their children, and many health professionals in primary care settings recommend them to their patients as a first‐line treatment (PAGB 2000). OTC medicines are available to the public from pharmacies, chemists and shops without medical or dental prescription in most countries, as opposed to prescription‐only medicines (POM). A national telephone survey of medication use in the US indicated that in a given week, 10% of children are given an OTC cough preparation by their carers (Vernacchio 2008). Numerous OTC cough preparations are available, but evidence regarding their efficacy is inconclusive. Some studies of cough preparations have been shown to reduce cough symptoms, whereas others found no effect compared with placebo (Banderali 1995; Freestone 1997; Kurth 1978; Smith 1993). In recent years there have been safety concerns that led to the withdrawal of OTC cough medicines containing antihistamines and antitussives in children under six years. These regulatory changes were based on rare fatalities identified by the Food and Drug Administration (FDA) in the US, highlighting the potential of these agents to cause harm, particularly if there is accidental overdosage. However, these medicines are still available for children over the age of six years. In addition, following concerns about QT prolongation and cardiac arrhythmias, the non‐narcotic antitussive clobutinol was withdrawn in the European Union in 2008.

Description of the intervention

Many studies have involved patients from different populations that have included participants with chronic cough due to underlying disease such as asthma or chronic obstructive pulmonary disease, or that were carried out on healthy volunteers in whom cough had been induced by chemical irritants (Gastpar 1984; Irwin 1993; Smith 1993). Other randomised controlled trials (RCTs) compared active agents and did not include a placebo. Cough preparations may contain different drugs with a variety of modes of action, which can make them difficult to compare (Morice 1998).

How the intervention might work

Non‐prescription oral OTC medicines for cough have different modes of action based on their active ingredients as follows.

  1. Antitussives, for example centrally acting opioid derivatives (Irwin 1993), or other peripherally active agents, act by reducing the cough reflex.

  2. Expectorants, i.e. drugs leading to increased bronchial mucus production, make secretions easier to remove by cough or ciliary transport (Ziment 1976).

  3. Mucolytics, i.e. drugs aiming to decrease the viscosity of bronchial secretions, act to make secretions easier to clear through coughing (Reynolds 1993).

  4. Antihistamine‐decongestant combinations, i.e. drugs that are combined antihistamine H1‐receptor antagonists and alpha‐adrenoceptor agonists, act by causing vasoconstriction of mucosal blood vessels thus reducing congestion (Morice 1998).

  5. Antihistamines, i.e. antihistamine H1‐receptor agonists, act by reducing histamine release and thus reducing local congestion and production of secretions.

  6. Other drug combinations, i.e. fixed drug combinations using different ingredients, have mechanisms of action based on individual ingredients.

  7. Honey is regarded as having bactericidal properties and there has been recent interest in its potential effectiveness in relieving the symptoms of URTI, including cough. In studies to date it has been administered as a single dose before bedtime or diluted in a non‐caffeinated beverage. An existing Cochrane review of honey for acute cough has concluded that there is no strong evidence for or against the use of honey in acute cough in children (Oduwole 2012).

Why it is important to do this review

Previous systematic reviews of OTC cough and cold preparations revealed that there is insufficient evidence for or against an effect of OTC cough preparations compared to placebo (Anonymous 1999; Smith 1993). However, these reviews did either not use a systematic search for RCTs (Anonymous 1999), or performed searches that were limited to the MEDLINE database (Smith 1993). By using a more extensive search strategy, this systematic review aims to answer the question of whether OTC medications used for the treatment of acute cough associated with URTI are effective.

Objectives

To assess the effects of oral OTC cough preparations for acute cough in children and adults in community settings.

Many different groups of OTC medicines are available, therefore we aimed to make comparisons only within groups of preparations with a similar mode of action or other similar features.

Methods

Criteria for considering studies for this review

Types of studies

All placebo‐controlled RCTs of oral OTC cough preparations for acute cough.

Types of participants

  1. Community or ambulatory settings in primary care and hospital outpatients.

  2. Children and adults with acute onset of cough (less than three weeks' duration).

We excluded studies testing OTC medicines for chronic cough (more than three weeks' duration), cough due to underlying respiratory disease (such as asthma, chronic obstructive pulmonary disease, pneumonia, tuberculosis, lung malignancy). We also excluded studies where cough was induced artificially (through inhalation of chemicals) in healthy volunteers.

Types of interventions

Non‐prescription oral OTC medicines for cough are classified according to their mode of action, as outlined above, and we have grouped them as follows.

  1. Antitussives, for example, centrally acting opioid derivatives (Irwin 1993).

  2. Expectorants, i.e. drugs leading to increased bronchial mucus production (Ziment 1976).

  3. Mucolytics, i.e. drugs aiming to decrease the viscosity of bronchial secretions (Reynolds 1993).

  4. Antihistamine‐decongestant combinations, i.e. drugs that are combined antihistamine H1‐receptor antagonists and alpha‐adrenoceptor agonists, which cause vasoconstriction of mucosal blood vessels (Morice 1998).

  5. Antihistamines, i.e. antihistamine H1‐receptor agonists.

  6. Other drug combinations, i.e. fixed drug combinations using different ingredients.

  7. Honey.

We excluded studies that used non‐oral preparations (for example, nasal sprays, inhalers, nebulised solutions) or that tested ingredients other than those accepted in Western (allopathic) medicine (for example, herbal or homeopathic medicines) because we felt that this review would have become too broad.

Types of outcome measures

Primary outcomes

All cough outcomes (such as frequency, severity, amount of sputum, improvement in cough symptoms using continuous and categorical data and different ways of measurement including cough sound pressure levels, cough counts, patient questionnaires, physician assessment, etc). We did not consider global patient or physician ratings of wellness or recovery as outcomes, unless these were directly related to cough symptoms.

Secondary outcomes

Significant adverse effects.

Search methods for identification of studies

Electronic searches

For this 2014 update we searched the Cochrane Central Register of Controlled Trials (CENTRAL) (2014, Issue 1) (accessed 26 March 2014), which contains the Cochrane Acute Respiratory Infections Group's Specialised Register; MEDLINE (January 2012 to March week 3, 2014); EMBASE (March 2012 to March 2014); CINAHL (2012 to March 2014); LILACS (2012 to March 2014) and Web of Science (2012 to March 2014). We used the search strategy in Appendix 1 to search MEDLINE and CENTRAL. We combined the MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity‐ and precision‐maximising version (2008 revision) Ovid format (Lefebvre 2011). We adapted the search string to search EMBASE (see Appendix 2), CINAHL (see Appendix 3), LILACS (see Appendix 4) and Web of Science (see Appendix 5).

For our previous 2012 update we searched the Cochrane Central Register of Controlled Trials (CENTRAL) (2012, Issue 3) (accessed 22 March 2012), which contains the Cochrane Acute Respiratory Infections Group's Specialised Register; MEDLINE (February 2010 to March week 1, 2012); EMBASE (March 2010 to March 2012); CINAHL (January 2010 to March 2012); LILACS (2010 to March 2012) and Web of Science (2010 to March 2012). Details of earlier searches are in Appendix 6.

Searching other resources

We searched personal collections of references and reference lists of articles, and wrote to authors of original studies, pharmaceutical companies and the Proprietary Association of Great Britain about information on unpublished studies. We did not impose any constraints based on language or publication status.

Data collection and analysis

Selection of studies

Two review authors (SS, TF) independently screened potentially relevant citations and applied the selection criteria using an in/out/pending sheet. We resolved any differences at any stage of the review by discussion.

Data extraction and management

Two review authors (SS, TF) independently extracted data and assessed the quality of studies. We contacted investigators for additional information if necessary and obtained translations of abstracts or papers if they were written in languages other than English or German. We resolved any disagreements by discussion.

Assessment of risk of bias in included studies

Since the 2010 update of this review we have adapted our original quality assessment and have used the 'Risk of bias' tool outlined in the Cochrane Handbook for Systematic Reviews of Interventions to assess the methodological quality of included studies (Higgins 2011). Two review authors (SS, TF) independently carried out these assessments. The domains considered are now described within the Characteristics of included studies table.

  1. Random sequence generation.

  2. Allocation concealment.

  3. Blinding of participants and personnel.

  4. Blinding of outcome assessment.

  5. Incomplete outcome data.

  6. Selective reporting.

  7. Other bias.

Measures of treatment effect

Due to the small numbers of trials in each category, the limited quantitative data available and the marked differences between trials in terms of participants, interventions and outcome measurement we felt that pooling of the results was inappropriate and we undertook no meta‐analysis. The effect of individual treatments is summarised as outlined in the original studies using mean differences in scores for continuous data where available or simple presentations of means in each group or comparison of proportions for dichotomous data.

Unit of analysis issues

All included studies were RCTs with randomisation occurring at the level of individual participants, so there was no indication to consider unit of analysis errors in this review.

Dealing with missing data

Due to the limited quantitative data available for this review, we presented simple descriptions of individual study outcomes within the pre‐specified grouping of different treatment groups. Issues relating to missing data and follow‐up are presented in the 'Risk of bias' sections in the Characteristics of included studies table.

Assessment of heterogeneity

The studies included in this review were clinically heterogeneous and provided limited data, so we undertook no meta‐analysis.

Assessment of reporting biases

There is no reason to suspect that publication bias affected the outcomes of this review. We conducted a comprehensive search of the literature with no language or publication restrictions. For the original review we also sought information from experts in the area including pharmaceutical companies and the Proprietary Association of Great Britain and Ireland. As we did not perform a meta‐analysis, we did not generate funnel plots.

Data synthesis

We undertook no meta‐analysis for this review.

Subgroup analysis and investigation of heterogeneity

Effects of treatment are presented within relevant treatment groups for both children and adults to allow comparison of related medications.

Sensitivity analysis

We undertook no meta‐analysis and limitations of the review are addressed within the Discussion section.

Results

Description of studies

Results of the search

Our initial search in 2001 identified 328 potentially relevant RCTs, which we screened for retrieval of paper copies. At that stage we excluded 235 abstracts for the following main single reasons: study not a RCT (n = 19 trials); study not placebo‐controlled (n = 39); study not testing an OTC cough medicine (n = 86); cough artificially induced (n = 26); or participants with chronic cough lasting more than three weeks (n = 65). We retrieved paper copies of 93 RCTs for more detailed evaluation. We excluded a further 72 trials because studies were not RCTs (n = 4); were not placebo‐controlled (n = 2); were not testing OTC cough medicines (n = 23); induced cough artificially (n = 3); included participants with chronic cough (n = 25); or did not report any cough outcomes (n = 15).

The search conducted for the update in 2004 identified 477 potential titles for screening and we identified three additional RCTs as eligible for inclusion, with two of these being different arms of a three‐arm RCT (Korppi 1991a; Korppi 1991b; Pavesi 2001).

The search conducted for the update in 2007 identified 799 potential titles for screening and we identified one additional RCT as eligible for inclusion (Paul 2004).

The search conducted for the 2010 update identified 112 potential titles for screening and we identified one additional RCT as eligible for inclusion (Mizoguchi 2007).

The search conducted for the update in 2012 identified 252 potential titles for screening and no additional eligible studies. We identified one potentially eligible study that we excluded as it had no placebo control group (Shadkam 2010).

The search conducted for this 2014 update identified 421 potential titles for screening and we identified three additional RCTs as eligible for inclusion (Albrecht 2012; Avner Cohen 2012; Bhattacharya 2012).

In total, 2389 potential titles have been identified for screening since the first review in 2001 and 29 RCTs were identified as eligible for inclusion in this review. See Figure 1.


Flow diagram.

Flow diagram.

Included studies

In this 2014 update we included 29 RCTs involving 4835 participants. Nineteen of these trials were in adults (n = 3799) and 10 in children (n = 1036). The Characteristics of included studies table contains data on the number of participants randomised to the interventions, age, sex, smoking status, study setting, definition of illness, drug dosage, frequency and duration of treatment, and outcome information. Most adult trials were in young adults with mean ages ranging from 23 to 48 years. Ages in studies in children ranged from six months to 18 years. Six trials were more than 20 years old. Nearly half of the studies (13 out of 29) were carried out in the US, with the remaining trials located in the UK (five), Finland (three), Germany (two), Italy (one), India (two), South Africa (one), Thailand (one) and Israel (one). The ages of participants ranged from six months to over 70 years. Most studies were different in their definition of illness, the content of the drug preparation, drug dosage, the frequency of doses and the treatment duration (ranging from a single dose to 18 days), making comparison of trials and quantitative analysis difficult. Table 1 provides an overview of the included studies, treatment types, treatment duration and results and highlights this variation.

Open in table viewer
Table 1. Overview of studies and outcomes

Group

Study ID

Number participants

Treatment and duration

Outcome and result

Notes

1. Antitussives

Antitussives

Adults

Adams

108

Moguisteine
3.5 days

Cough score (10‐point scale)
MD 0.9 on day 4

NS

Eccles

81

Codeine

4 days

16‐item cough severity score
MD 1.6

NS

Freestone

82

Codeine

Single dose

5‐point symptom rating scale
MD 1.0

NS

Lee

44

Dextromethorphan

Single dose

Cough severity score (3‐point scale)
MD at 180 minutes 0.5

NS

Parvez

451

Dextromethorphan

Single dose

Cough recordings at 180 minutes

Significant reduction in cough counts with treatment

Pavesi

710

Dextromethorphan

Single dose

Cough recordings at 180 minutes

Significant reduction in various cough measures with treatment

Antitussives

Children

Bhattacharya

120

Dextromethorphan
3 days

Composite 5‐item symptom score
MD 0.4

NS

Korppi

50

Dextromethorphan

3 days

4‐item cough symptoms score day 3
MD 0.04

NS

Paul

100

Dextromethorphan

Single dose

Composite 5‐item symptom score
MD 0.79

NS

Taylor

57

Dextromethorphan or codeine (3 arms)

3 nights

4‐item symptom score

NS

2. Expectorants

Expectorants

Adults

Albrecht

378

Guaifenesin 1600 mg bd
7 days

8‐item composite symptom score, day 4
MD 2.6, P value = 0.04

NS difference at day 7, reported day 4 outcomes

Kuhn

65

Guaifenesin 480 mg qid
30 hours

Cough recordings and patient‐rated improvement in cough frequency and severity

NS

Robinson

239

Guaifenesin 200 mg qid
3 days

Patient rating as 'helpful'
Int 75% versus Con 31%

P value < 0.01

Expectorants

Children

No studies

3. Mucolytics

Mucolytics

Adults

Nesswetha

99

Bromhexine

4 days

Cough frequency (cough every 2 to 4 minutes)

Int 8.6% versus Con 15.2%

P value < 0.02

Mucolytics

Children

Nespoli

40

Letosteine
10 days

Composite symptom score
Average 2‐point difference between intervention and control favouring intervention day 4 to 10

P value < 0.01

4. Antihistamine/decongestants

Antihistamine/

decongestants

Adults

Berkowitz

283

Loratadine and pseudoephedrine
5 days

3‐item cough score
MD 0.2

NS

Curley

73

Dexbrompheniramine and pseudoephedrine

7 days

4‐item cough score
MD 0.6

P value < 0.05

Antihistamine/

decongestants

Children

Clemens

59

Brompheniramine and phenylpropanolamine

2 days

7‐item cough scores
MD 0.1

NS

Hutton

96

Brompheniramine and phenylephrine and propanolamine

2 days

9‐item symptom score, % reported as improved
Int 67% versus Con 58%

NS

5. Other combinations

Other combinations

Adults

Kurth

113

Dextromethorphan plus salbutamol

3 days

% improved day 3
Int 45% versus Con 27%

P value = 0.05

Mizoguchi

485

Combination syrup
Single dose

Cough score

MD 0.42, day 2

P value < 0.0001

Thackray

70

'Vicks'
2 days, cross‐over after 24 hours

Patient preference

Int 66% versus Con 27%

P value < 0.01

Tukiainen

108

Dextromethorphan with and without salbutamol

4 days

5‐item symptom score

MD 0.11

NS

Other combinations

Children

Korppi

51

Dextromethorphan plus salbutamol

3 days

Cough symptoms

MD 0.12, day 3

NS

Reece

43

Combination syrups

Satisfactory antitussive response
Int both 69% versus Con 57%

NS

6. Antihistamines

Antihistamines

Adults

Berkowitz

91

Terfenadine

4 to 5 days

3‐item symptom score

MD 0.2

NS

Gaffey

250

Terfenadine

3.5 days

Symptom score (scores not reported)

NS

Antihistamines

Children

Bhattacharya

120

Promethazine

3 days

5‐item symptom score

MD 0.3

NS

Paul

100

Diphenhydramine

Single dose

5‐item symptom score
MD 1.73

NS

Sakchainanont

143

Clemastine and chlorpheniramine

3 days

4‐item symptom score, % reporting improvement
Int 40% versus Con 28%

NS

7. Honey

Honey

Adults

No studies

Honey Children

Avner‐Cohen

300

3 types honey (4 arms)
Single dose

5‐item composite symptom score

Significant symptom reductions in all honey groups compared to controls

Con: control
Int: intervention
MD: mean difference
NS: not significant
qid: 4 times a day

Excluded studies

The commonest reasons for excluding studies were lack of a placebo control, that cough was artificially induced or lasted longer than three weeks, or cough outcomes were not clearly reported. See Characteristics of excluded studies table.

Risk of bias in included studies

These are summarised in Figure 2. All studies were randomised controlled trials but methods for allocation concealment and random sequence generation were reported in only five of the 29 included studies. All studies were placebo‐controlled based on inclusion criteria and the majority reported blinding of providers (24 of 29 studies), but only a minority reported on blinding of outcome assessors (9 of 29 studies). There was minimal risk of attrition bias. There was potential risk of bias related to funding sources as studies funded by pharmaceutical companies or other providers (12 of 29 studies) were more likely to have positive results.


'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Most studies did not report sufficient details on randomisation and allocation schedules to make meaningful conclusions about the potential for selection bias. Only five of the 29 trial reports stated the randomisation process, which was adequate in three trials.

Loss to follow‐up was well documented in 20 studies with differential loss to follow‐up in the treatment arms reported in five studies, and with the potential for attrition bias difficult to assess for the remaining studies. Only three of the studies fulfilled all the quality criteria. Only eight trials reported a power calculation.

Blinding

All studies were placebo‐controlled thus patients were blinded. In 24 of the 29 included studies providers were also blinded. Only nine of the 29 included studies reported whether outcome assessors had been blinded, which could lead to potential detection and performance bias.

Incomplete outcome data

In general there was adequate follow‐up of participants with 19 of the 29 studies reporting adequate follow‐up and the remainder being unclear due to lack of reporting. A number of studies dated back many years, therefore it was often impossible to obtain additional trial data. As the reporting of potential causes of bias was poor in many trials, we did not formally examine the trial efficacy versus the trial quality and therefore only summarised the available data in the 'Risk of bias' section of the Characteristics of included studies. These contain summary data on randomisation processes used, blinding to treatment allocation, drop‐outs/losses to follow‐up and any additional comments.

Other potential sources of bias

Twelve of the 29 included studies were fully or partly supported by pharmaceutical or other supplier companies, which provided grants, supplied the drugs in question or gave assistance with the study (Adams 1993; Avner Cohen 2012; Berkowitz 1991; Gaffey 1988; Mizoguchi 2007; Parvez 1996; Pavesi 2001; Reece 1966; Robinson 1977; Sakchainanont 1990; Thackray 1978; Tukiainen 1986). Seven out of the 12 studies supported by industry showed positive results compared to nine out of 18 trials where no support was reported.

Effects of interventions

We grouped the trials according to drug class into antitussives, expectorants, mucolytics, antihistamine‐decongestant combinations, other combinations and antihistamines. The number of studies in each group ranged from one to a maximum of six. Cough outcomes included frequency, severity and night‐time symptoms and were measured in many different ways, for example, participant self report by symptom scores (interviews, questionnaires, diaries), physician assessment, observation by parents, cough sound pressure levels obtained by recordings via a microphone and tape recordings. Nineteen studies out of 29 reported data on adverse effects and six studies reported data on compliance with medication. Eleven out of the 29 trials reported quantitative data for the cough that could potentially have been used for meta‐analysis. Due to the small numbers of trials in each category, the limited quantitative data available and the marked differences between trials in terms of participants, interventions and outcome measurement, we felt that pooling of the results was inappropriate.

1. Antitussives

1.1 Studies in adults

We included six trials involving 1526 participants that compared antitussives with placebo.

Codeine was tested in two trials and appeared no more effective than placebo in reducing cough symptoms (Eccles 1992; Freestone 1997). One of these studies (n = 81) tested codeine in a two‐phase study (laboratory and home) at a dose of 30 mg four times daily for four days (Eccles 1992), and codeine was no more effective than placebo either as a single dose or in a total daily dose of 120 mg, reported on a five‐point cough severity score (cough severity scores were expressed as area under the curve for eight measures over five days with intervention group 17.2 versus control group 18.8, P value = 0.23). The second study of codeine (n = 82) only tested the effect of a single 50 mg dose (Freestone 1997), and cough was assessed via microphone using cough sound pressure levels 90 minutes after drug administration, cough frequency counts and subjective scores. The mean subjective score on a five‐point rating scale was reduced from 2.0 to 1.0, 90 minutes after treatment in both treatment groups (P value = 0.8). Neither study provided any data on side effects.

Dextromethorphan was tested in three of the included studies (Lee 2000; Parvez 1996; Pavesi 2001). One report on a series of three successive studies on a total of 451 adults favoured dextromethorphan 30 mg given in a single dose to placebo in terms of cough counts (measured through cough acoustic signals using a microphone on the nose) and subjective visual analogue scales (Parvez 1996). Differences in mean changes of cough counts between active treatment and placebo varied from 19% to 36% (P value < 0.05) in the three studies (up to a net difference of eight to 10 coughing bouts every 30 minutes). This study did not report on side effects.

A study involving 44 participants tested a single 30 mg dose of dextromethorphan versus placebo (Lee 2000). Both treatment groups showed a decline in cough frequency (from 50 to 19 per 10‐minute period in the active treatment arm compared with 42 to 20.5 in the placebo arm, P value = 0.38 at 180 minutes follow‐up). Mean subjective cough scores showed a decline from 2.0 to 1.0 in the active treatment group compared to a decline from 2.0 to 1.5 in the placebo group (mean difference in decline in cough scores 0.5 at 180 minutes, P value = 0.08).

Pavesi and colleagues also tested a single 30 mg dose of dextromethorphan versus placebo (Pavesi 2001). Outcomes were measured through a three‐hour continuous cough recording, measuring cough bouts, cough components, cough effort, cough intensity and cough latency. Average treatment difference was 12% to 17% in favour of dextromethorphan for cough bouts (P value = 0.004), cough components (P value = 0.003) and cough effort (P value = 0.001), with an increase in cough latency (P value = 0.002).

One trial in 108 adults comparing moguisteine at a total daily dose of 600 mg for three and a half days with placebo showed no difference apart from cough reduction in individuals with more severe night cough (Adams 1993). The overall mean cough score difference was 0.9 (intervention group score 6.4 versus control group score 5.5). There were no differences in the proportion rated as being improved by the investigators (intervention 42/48 versus control 41/48). There were more side effects in the treatment group (22%) compared to placebo (8%), which mainly included nausea, vomiting and abdominal pain. There were four withdrawals in the treatment group due to adverse effects.

1.2 Studies in children

We included four trials involving 327 participants that compared antitussives with placebo in children.

One study involving 57 children with night cough compared a single dose for three nights of dextromethorphan or codeine with placebo (Taylor 1993). Mean cough and composite scores decreased in each of the three treatment groups on each day of the study. Neither dextromethorphan (cough score reduction of 2.1, P value = 0.41) nor codeine (cough score reduction of 2.2, P value = 0.70) was more effective than placebo (cough score reduction of 2.2) on day three.

Another study involving 50 children compared dextromethorphan 1.5 mg per ml 5 ml three times a day for children under seven years and 10 ml three times daily for older children with placebo (Korppi 1991a). There were no differences between the groups in terms of parent‐recorded symptom scores (mean difference in cough symptom scores on day 3 of 0.04) or adverse effects, which were generally mild.

A third study involving 100 children compared a single nocturnal dose of dextromethorphan (dose based on child's age: age two to five, 7.5 mg; age six to 11, 15 mg; age 12 to 18, 30 mg with either a single dose of an antihistamine or with placebo) (Paul 2004). Dextromethorphan was no more effective than diphenhydramine or placebo in reducing cough frequency or impact on child or parental sleep (composite symptom score intervention 10.06 versus control 10.85, mean difference 0.79).

A new study included an arm where 40 children were given 5 mg dextromethorphan six‐ to eight‐hourly for three days and found no difference from placebo treatment (Bhattacharya 2012). There was no significant difference in composite symptom scores on day three (intervention 4.6 versus control 5.0, mean difference 0.4). Adverse events were reported in 34% of participants in the dextromethorphan group and 32% of participants in the antihistamine group compared to 5% of participants in the placebo group.

2. Expectorants

2.1 Studies in adults

Three trials with a total of 604 participants compared guaifenesin with placebo (Albrecht 2012; Kuhn 1982; Robinson 1977). In one study (n = 239), 75% of participants taking guaifenesin stated that the medicine was helpful in terms of reducing cough frequency and intensity compared to 31% in the control group (P value < 0.01) at 72 hours (Robinson 1977). Four participants (two in each group) reported side effects including nausea and hives in the active treatment group and headaches, drowsiness and excessive perspiration in the placebo group.

The second study (n = 65) evaluated an antitussive rather than an expectorant effect of guaifenesin, which is usually classified as an expectorant (Kuhn 1982). Individuals in both groups reported improvement with respect to cough frequency (100% in the active group versus 94% for placebo, P value = 0.5) and cough severity (100% in the active treatment group versus 91% in the placebo group, P value = 0.2) at 36 hours. Guaifenesin reduced sputum thickness significantly in 96% of participants compared to 54% in the placebo group (P value = 0.001). This study allowed aspirin and paracetamol for participants after inclusion in the study, and the vehicle contained 95% alcohol. Adverse effects were not reported on.

A third new study compared extended‐release guaifenesin with placebo in 378 participants aged greater than 12 years over seven days and found no difference in total spontaneous symptom severity scores on day seven but reported day four results in detail, which indicated a reduction in mean score from baseline of 7.1 in the intervention group compared to a reduction of 5.7 in the control group (mean difference (MD) 2.6, P value = 0.04) (Albrecht 2012).

2.2 Studies in children

We did not include any studies that tested expectorants in children, partly because none of the outcomes under study were reported on.

3. Mucolytics

3.1. Studies in adults

One trial involving 99 participants compared bromhexine 5 mg three times daily for an average of four days with placebo (Nesswetha 1967). Frequent cough (every two to five minutes) was more prevalent in the placebo group (15.2%) compared to active treatment (8.6%, P value < 0.02) leading to a risk ratio reduction of about 50% for frequent cough. This study did not report on any adverse effects.

3.2 Studies in children

One trial involving 40 children compared the mucolytic letosteine (preparation not available in the UK and other parts of the world) at a dose of 25 mg three times daily for 10 days with placebo (Nespoli 1989). The symptom score on a four‐point scale favoured active treatment from day four until day 10, with an average difference of about 0.2 points (P value < 0.01). No adverse effects were reported in either group.

4. Antihistamine‐decongestant combinations

4.1 Studies in adults

Two trials in adults with a total of 356 participants compared antihistamine‐decongestant combinations with placebo (Berkowitz 1989; Curley 1988). One trial comparing loratadine/pseudoephedrine (5 mg/120 mg twice daily for four days) with placebo (n = 283) did not show statistically significant differences in cough scores reported in patient diaries between both groups (Berkowitz 1989). Thirty per cent of participants in the active treatment group reported adverse effects including dry mouth, headache and insomnia compared to 21% in the control group.

The second trial (n = 73) compared dexbrompheniramine/pseudoephedrine (6 mg/120 mg twice daily for one week) with placebo. The mean severity rank of cough on a scale from zero to four obtained through a patient diary was less in the active treatment group (1.4) than in the placebo group (2.0) on days three to five (P value < 0.05) (Curley 1988). There was an increased severity of dizziness and dry mouth in the active drug group on days five to seven, and two to 10, respectively (exact figures not reported, P value = or < 0.01).

4.2 Studies in children

Two studies involving 155 children compared antihistamine‐decongestant combinations with placebo (Clemens 1997; Hutton 1991).

One study including 59 participants found that brompheniramine/phenylpropanolamine (2 mg/12.5 mg, half the dose for children from six months to one year, on a four‐hourly 'as needed' basis for 48 hours) was no more effective than placebo in reducing the number of children coughing two hours after each dose (49.0% versus 43.1%, P value = 0.66). A higher proportion of children was reported to be asleep in the active treatment group (46.6%) than in the placebo group (26.5%, P value = 0.53), and no other adverse effects were reported (Clemens 1997).

In the second study (n = 96), a combination of brompheniramine/phenylephrine/propanolamine (see Characteristics of included studies table for full dosage details) led to a non‐statistically significant improvement in cough in 67% of children (reported by their parents) compared to 58% in the placebo group and 70% in the group receiving no treatment (Hutton 1991). Side effects were rare and included one child with loose stools in the placebo group and one child reported to be hyperactive in the active drug group. A second child in the drug group was reported to be sleepier than usual.

5. Other drug combinations

For the constituent ingredients of the drug combination formulations included in the review please refer to the Characteristics of included studies table.

5.1 Studies in adults

Four studies involving 836 people compared other combinations with placebo (Kurth 1978; Mizoguchi 2007; Thackray 1978; Tukiainen 1986). These studies were very heterogeneous and used very different drug preparations and dose frequency, limiting their comparability.

In one trial (n = 113) EM‐Vier (Minetten) given six times daily was more effective in reducing coughing fits (25% versus 11%, P value < 0.01) and the urge to cough (27% versus 14%, P value < 0.01) compared to placebo in the first seven days (Kurth 1978). There were no adverse effects in either group.

In a trial of Vicks Medinite syrup (n = 70) at a single dose at bedtime for two days, 57.6% of participants in the active treatment group rated the formulation as "good" or better in relieving cough compared to 32.2% in the placebo group (P value < 0.01) (Thackray 1978). Seven participants in the active treatment group reported giddiness/drowsiness compared to four participants in the placebo group.

Another study (n = 108) compared a dextromethorphan/salbutamol combination and dextromethorphan alone with placebo (Tukiainen 1986). There was spontaneous improvement of cough in all groups, and there were no statistically significant differences in cough scores between active treatments and placebo for both cough frequency and severity during the day. Dextromethorphan/salbutamol was superior to placebo or dextromethorphan alone in relieving cough at night (mean symptom score 0.19 versus 0.67 and 0.44, respectively on day four, P value < 0.01). The dextromethorphan/salbutamol combination led to more tremor than placebo (no figures given, P value < 0.05), and no serious adverse effects were reported.

A further study (n = 485), identified for the 2009 update of this review, compared a single nocturnal dose of a compound containing four agents each with potential to deal with the different symptoms of the common cold, i.e. paracetamol plus dextromethorphan plus doxylamine plus ephedrine (Mizoguchi 2007). We only report the cough‐related outcomes. The outcomes in this study were measured over the following two days and included proportions who reported improvements in cough three hours after taking the treatment and mean cough scores on day one and day two. There was a significant improvement in mean cough score the morning after treatment and the following day (mean cough score 2.5 versus 2.08 on day two, MD 0.42, P value < 0.0001). There were also improvements in the proportion reporting improvement in cough three hours after taking the medication (intervention 57% and control 43%). There were 19 adverse events in the study in 14 patients with no difference between treatment and control. However, there was one serious adverse event described as a severe episode of somnolence in the active treatment group.

5.2 Studies in children

We included two trials with 94 participants examining other drug combinations.

One trial involving 43 children tested two paediatric cough syrups (Triaminicol syrup and Dorcol paediatric cough syrup) (Reece 1966). Compared to placebo, 69% of children in both active treatment groups showed a satisfactory response reported by their parents compared to 57% of children in the placebo group, which did not reach statistical significance (P value = 0.5). Adverse effects were not reported.

One trial in 51 children compared a combination of dextromethorphan 1.5 mg per ml and salbutamol 0.2 mg per ml 5 ml three times daily for children under the age of seven or 10 ml three times a day for older children with placebo (Korppi 1991b). There were no differences between the groups in terms of parent‐recorded symptom scores or adverse effects, which were generally mild.

6. Antihistamines

6.1 Studies in adults

Two trials involving 350 adult participants compared antihistamines with placebo (Berkowitz 1991; Gaffey 1988). Antihistamines were no more effective than placebo in relieving cough symptoms. Terfenadine was tested in two studies. In one of these studies (n = 100), terfenadine at a dose of 120 mg twice daily for four to five days led to a mean cough score (measured by physicians' evaluation on a scale from zero to three with higher scores meaning more coughing) of 0.8 in the active treatment group compared to 0.65 in the placebo group (MD 0.15, P value = 0.35) (Berkowitz 1991). Possible adverse effects were rare in both groups, with headache being the most common complaint (6.1% of participants in the active treatment group compared to 4% in the placebo group).

The second study (n = 250) tested terfenadine at a dose of 60 mg twice daily for three and a half days (Gaffey 1988). There were no statistically significant differences in self reported symptoms scores for cough (exact figures not reported) between groups. Side effects were uncommon in both treatment groups, with the most common complaint being excess fatigue in 12% of participants receiving terfenadine compared to 10% in the placebo group.

6.2 Studies in children

We included three trials involving 363 children comparing antihistamines with placebo.

One study compared the antihistamines clemastine (0.05 mg/kg/day) and chlorpheniramine (0.35 mg/kg/day) for three days with placebo (Sakchainanont 1990). There was spontaneous improvement in all groups. In both active treatment groups, cough scores observed by physicians and participants improved in 39.6% of individuals compared with 27.6% in the placebo group, which did not reach statistical significance (P value = 0.2). Drowsiness and sleepiness were reported in 20% of children with no difference between the groups.

The second trial included an arm in which children received diphenhydramine in a single nocturnal dose and were compared with children receiving placebo (Paul 2004). Diphenhydramine was no more effective than dextromethorphan or placebo in reducing cough frequency or impact on child or parental sleep.

The third study, added in the current update, had one arm with 40 children who received promethazine at a dose of 0.5mg/kg/dose eight‐hourly (Bhattacharya 2012). There was no difference from placebo in cough and sleep‐related outcomes. Thirteen children receiving promethazine experienced adverse effects compared to two children receiving placebo.

7 Honey

One new study compared two days of treatment with three types of honey with placebo in 300 children aged one to five years (Avner Cohen 2012). Children were administered 10 g of honey (eucalyptus honey, citrus honey or labetiae honey) or silan date extract as placebo, all provided in identical containers. Parents were told to give it as a single dose or to dilute it in a non‐caffeinated beverage 30 minutes before the child was due to go to sleep. Children receiving honey had significant improvements in total symptom scores compared to placebo with no significant differences in adverse effects between the honey and placebo groups.

Adverse effects

Twenty‐one of the 29 included studies reported on adverse effects. There was a wide variation in the range of reported adverse effects across studies from less than 1% in some studies to approximately 30% in others. No study reported adverse effects as being serious. The adverse effects were generally higher in intervention participants compared to those taking placebo and this was particularly the case for preparations containing antihistamines and dextromethorphan.

Discussion

Summary of main results

We found no good evidence for or against the effectiveness of over‐the‐counter (OTC) medications in acute cough, which confirms the findings of two previous reviews (Anonymous 1999; Smith 1993). The number of trials in each group of drugs was small, there was poor overall quality of the studies and studies showed conflicting evidence. In total, 11 of the 29 included trials showed a positive result, whereas 18 did not show active treatment to be superior to placebo. Eight out of the 12 studies that were supported by the pharmaceutical industry showed positive results compared to four positive studies out of the 15 trials that did not report any conflict of interest. The results of trials did not appear to be related to their sample size or length of follow‐up. We did not formally examine the trial efficacy versus trial quality because of the lack of reported data.

Overall completeness and applicability of evidence

The results of this systematic review have to be interpreted with caution as the number of trials in each group was small. There were marked differences between the studies even within groups of drugs with similar mode of action, making it difficult to compare trials directly. In addition, there is variation between countries in relation to medications available OTC, making international comparisons more difficult. Inclusion and exclusion criteria for participants varied, and active drugs were administered in different total daily doses. The duration of drug therapy varied from a single‐dose treatment to an 18‐day course. For example, six studies testing antitussives, either alone or in combination with other agents, used short‐term cough relief after a single dose as an outcome (Freestone 1997; Lee 2000; Mizoguchi 2007; Parvez 1996; Paul 2004; Pavesi 2001), whereas more relevant outcomes for patients would be the effect after one day, three days or a week. Outcomes were assessed and measured in many different ways, which included questionnaires, cough severity scores, acoustic signals, tape recordings, daily diaries and assessment by a physician.

Most studies failed to provide quantitative data on cough as our main outcome of interest, which made it very difficult to assess whether positive study results were clinically relevant. Quantitative data that could be combined showed wide confidence intervals, although there was no evidence of statistical heterogeneity. Many included studies failed to report adverse effects adequately, and patient compliance with the treatment was not discussed in the vast majority of study reports. Four studies carried out multiple comparisons, thereby increasing the probability of a type I error (Albrecht 2012; Berkowitz 1989; Parvez 1996; Pavesi 2001).

Quality of the evidence

Most studies failed to provide quantitative data on cough. As this was our main outcome of interest, it was very difficult to assess whether positive study results were clinically relevant. Quantitative data that could be combined showed wide confidence intervals, although there was no evidence of statistical heterogeneity. Many included studies failed to report adverse effects adequately and patient compliance with the treatment was not discussed in the majority of study reports. The overall quality of trials is dubious and there are conflicting results between trials in each medication group. The method of outcome measurement and the resulting magnitude of effect were unclear or not very well reported in some studies.

Potential biases in the review process

Eleven of the 26 included studies were funded by the pharmaceutical industry as outlined in the 'Risk of bias' section in the Results. Studies funded in this way were more likely to report positive results. However, despite this potential bias the review does not provide evidence of the effectiveness of OTC cough medicines for acute cough.

Agreements and disagreements with other studies or reviews

The findings of this review and other related published evidence were considered by an expert panel of the US Food and Drug Administration (FDA) in October 2007 and there was consensus that there is limited evidence to support the recommendation to use OTC cough medicines for acute cough in children (FDA 2007). The review findings are also supported by a recent non‐Cochrane systematic review, which found few studies that examined the effectiveness of diphenhydramine for acute cough despite its widespread use, and these studies indicated limited clinical effectiveness (Bjórnsdóttir 2001).

Flow diagram.
Figures and Tables -
Figure 1

Flow diagram.

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.
Figures and Tables -
Figure 2

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Table 1. Overview of studies and outcomes

Group

Study ID

Number participants

Treatment and duration

Outcome and result

Notes

1. Antitussives

Antitussives

Adults

Adams

108

Moguisteine
3.5 days

Cough score (10‐point scale)
MD 0.9 on day 4

NS

Eccles

81

Codeine

4 days

16‐item cough severity score
MD 1.6

NS

Freestone

82

Codeine

Single dose

5‐point symptom rating scale
MD 1.0

NS

Lee

44

Dextromethorphan

Single dose

Cough severity score (3‐point scale)
MD at 180 minutes 0.5

NS

Parvez

451

Dextromethorphan

Single dose

Cough recordings at 180 minutes

Significant reduction in cough counts with treatment

Pavesi

710

Dextromethorphan

Single dose

Cough recordings at 180 minutes

Significant reduction in various cough measures with treatment

Antitussives

Children

Bhattacharya

120

Dextromethorphan
3 days

Composite 5‐item symptom score
MD 0.4

NS

Korppi

50

Dextromethorphan

3 days

4‐item cough symptoms score day 3
MD 0.04

NS

Paul

100

Dextromethorphan

Single dose

Composite 5‐item symptom score
MD 0.79

NS

Taylor

57

Dextromethorphan or codeine (3 arms)

3 nights

4‐item symptom score

NS

2. Expectorants

Expectorants

Adults

Albrecht

378

Guaifenesin 1600 mg bd
7 days

8‐item composite symptom score, day 4
MD 2.6, P value = 0.04

NS difference at day 7, reported day 4 outcomes

Kuhn

65

Guaifenesin 480 mg qid
30 hours

Cough recordings and patient‐rated improvement in cough frequency and severity

NS

Robinson

239

Guaifenesin 200 mg qid
3 days

Patient rating as 'helpful'
Int 75% versus Con 31%

P value < 0.01

Expectorants

Children

No studies

3. Mucolytics

Mucolytics

Adults

Nesswetha

99

Bromhexine

4 days

Cough frequency (cough every 2 to 4 minutes)

Int 8.6% versus Con 15.2%

P value < 0.02

Mucolytics

Children

Nespoli

40

Letosteine
10 days

Composite symptom score
Average 2‐point difference between intervention and control favouring intervention day 4 to 10

P value < 0.01

4. Antihistamine/decongestants

Antihistamine/

decongestants

Adults

Berkowitz

283

Loratadine and pseudoephedrine
5 days

3‐item cough score
MD 0.2

NS

Curley

73

Dexbrompheniramine and pseudoephedrine

7 days

4‐item cough score
MD 0.6

P value < 0.05

Antihistamine/

decongestants

Children

Clemens

59

Brompheniramine and phenylpropanolamine

2 days

7‐item cough scores
MD 0.1

NS

Hutton

96

Brompheniramine and phenylephrine and propanolamine

2 days

9‐item symptom score, % reported as improved
Int 67% versus Con 58%

NS

5. Other combinations

Other combinations

Adults

Kurth

113

Dextromethorphan plus salbutamol

3 days

% improved day 3
Int 45% versus Con 27%

P value = 0.05

Mizoguchi

485

Combination syrup
Single dose

Cough score

MD 0.42, day 2

P value < 0.0001

Thackray

70

'Vicks'
2 days, cross‐over after 24 hours

Patient preference

Int 66% versus Con 27%

P value < 0.01

Tukiainen

108

Dextromethorphan with and without salbutamol

4 days

5‐item symptom score

MD 0.11

NS

Other combinations

Children

Korppi

51

Dextromethorphan plus salbutamol

3 days

Cough symptoms

MD 0.12, day 3

NS

Reece

43

Combination syrups

Satisfactory antitussive response
Int both 69% versus Con 57%

NS

6. Antihistamines

Antihistamines

Adults

Berkowitz

91

Terfenadine

4 to 5 days

3‐item symptom score

MD 0.2

NS

Gaffey

250

Terfenadine

3.5 days

Symptom score (scores not reported)

NS

Antihistamines

Children

Bhattacharya

120

Promethazine

3 days

5‐item symptom score

MD 0.3

NS

Paul

100

Diphenhydramine

Single dose

5‐item symptom score
MD 1.73

NS

Sakchainanont

143

Clemastine and chlorpheniramine

3 days

4‐item symptom score, % reporting improvement
Int 40% versus Con 28%

NS

7. Honey

Honey

Adults

No studies

Honey Children

Avner‐Cohen

300

3 types honey (4 arms)
Single dose

5‐item composite symptom score

Significant symptom reductions in all honey groups compared to controls

Con: control
Int: intervention
MD: mean difference
NS: not significant
qid: 4 times a day

Figures and Tables -
Table 1. Overview of studies and outcomes