Introduction
Common cold (CC) is a viral infection of the upper airways also referred to as upper
respiratory tract infection (URI) and including acute rhinosinusitis, pharyngitis
and laryngitis. URI usually appears jointly with acute bronchitis (AB) and differentiation
is almost unfeasible [1]. Cough, wheezing and sometimes sputum production are the leading symptoms of AB,
but considerable overlap exists between these two conditions where symptoms are related
to different areas of the upper and lower airways [1].
CC/AB leads to approximately 75 to 100 million physician visits annually only in the
U.S. [2]. CC and other acute respiratory infections account for about 30 % of absenteeism
from school and approximately 40 % of time lost from work, including parents staying
home to care for their sick children and millions of employees suffering from a cold
[3]. CC/AB affects about 5 % of adults annually [4] and is one of the most common outpatient illnesses recorded by physicians; in the
UK, 30 to 40 % of registered patients consult their physicians for treatment of respiratory
tract infections each year [5]. Thus, although they are usually self-limiting, these clinical conditions pose a
very large financial burden for societies and a great symptom burden for patients.
Although the great majority of cases of CC/AB relates to viruses (mainly rhinovirus
(50 % of cases), influenza A and B viruses, parainfluenza virus, RSV, coronavirus,
adenovirus) and only up to 10 % of cases are caused by bacteria (mainly Mycoplasma pneumoniae, Chlamydia pneumoniae, Bordetella pertussis), antibiotics are the most common prescribed treatment worldwide [4]
[6]. But superiority of appropriate antibiotic treatment in otherwise healthy patients
could not be shown even in cases of bacterial infection [4]
[7]
[8]. According to recent Cochrane reviews, antibiotics should not be prescribed in uncomplicated
AB [8] and rhinosinusitis [9]; for persistent cough, evidence even in children is insufficient and recommendations
for antibiotic treatment cannot be given [10]. Overall, the small beneficial effect of antibiotic treatment in meta-analysis should
be considered in the context of potential adverse events, the cost of antibiotics
and, most important, of increasing antibiotic resistances. Furthermore, the usual
over-the-counter (OTC) treatment with mucolytic and antitussive agents shows no evidence
for or against their efficacy, i. e. is not supported by data from clinical trials
[4].
Assessment of pharmacotherapy
Assessment of pharmacotherapy
Proof of efficacy in contemporary pharmacotherapy is based on randomized controlled
trials (RCTs). Designing RCTs in CC/AB is especially demanding, since it is a complex,
self-limiting disease. There is a large variance in symptoms in CC, where patients
suffer from sore throat, rhinorrhea, nasal stuffiness, headache, cough, sputum production,
chest pain, and wheeze. All symptoms are highly variable, i. e. they may change spontaneously
during the course of the disease and in some cases decrease rapidly. For example,
using the Leicester Cough Monitor, an objective and valid tool to assess cough frequency
[11]
[12], Lee et al. found more than 50 % cough reduction after the first 24 hours in patients
with acute cough receiving no treatment [13]. Thus, high placebo effects must be taken into account when symptomatic treatment
options are evaluated. On the other hand, cough is the most bothersome and long-lasting
symptom in CC [14], with a mean resolution time as long as 25 days.
To examine the effects of pharmacotherapy in complex diseases, composite assessment
scores are commonly used. Rheumatologists employ different composite measures to assess
rheumatoid arthritis [15], psychiatrists commonly rate depression in clinical research by multiple item questionnaires
such as the Hamilton Depression Rating Scale (HDRS) [16] or the Montgomery-Asberg Depression Rating Scale (MADRS) [17], and physicians treating athletes experienced concussion have been relying on the
SCAT2-score (Sport Concussion Assessment Tool) for years [18]. Potential benefits of composite outcome scales are their ability to measure clinically
relevant changes even in cases where not all of the single endpoints of interest changed
or where the endpoints changed to a different extent. Patient-reported outcome instruments
are also valuable for evaluating symptom improvements and are used in different areas
of respiratory medicine [19], e. g. chronic obstructive pulmonary disease (COPD), asthma research, pulmonary
arterial hypertension (PAH), etc.
The need for a reliable instrument to estimate disease severity in AB is supported
by the efforts of Mwachari and colleagues who employed a multiple item instrument
called the Acute Bronchitis Severity Score (ABSS) in 2007 [20]
[21]. Patients rate the degree of five items (overall severity of health, day cough,
night cough, limited daily activity, and fever) bothering them within the past 24
hours. However, “fever” is rather unusual in CC/AB and apart from the very initial
phase it cannot be considered as an appropriate item to rate the severity. Therefore,
its use as a score item should be seen rather critically. The main disadvantage of
this patient-rated questionnaire is that the ABSS is only validated in a high human
immunodeficiency virus (HIV) prevalence setting fpr patients aged 18 years or older
[20]. It was adopted in only one clinical trial in Africa [21].
The Bronchitis Severity Scale (BSS)
The Bronchitis Severity Scale (BSS)
Concerning AB sensu stricto, Dome et al. filled the gap of an appropriate outcome
measure by introducing the Bronchitis Severity Scale (BSS) in 1996 [22]. The scale comprises the five most important features of AB, i. e. cough, sputum
production (expectoration), rales/rhonchi (auscultation), chest pain during coughing,
and dyspnoea [23]. The BSS is an instrument which combines objective and subjective items, because
the assessment is based on the investigator’s clinical evaluation in conjunction with
the subjective feedback of the patient. Each constituent of the BSS is assessed by
the investigator using a 5-point verbal rating scale ranging from 0 to 4 (0: absent;
1: mild; 2: moderate; 3: severe; 4: very severe). The total score of BSS is the sum
of the five ratings with a maximum of 20 points.
Since its introduction in 1996 [22], the BSS has been successfully used in many clinical studies as a main outcome measure
in patients suffering from AB. Most of these research groups evaluated the efficacy
of a herbal drug preparation (EPs 7630[1]) extracted from the roots of Pelargonium sidoides [24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]. EPs 7630 has been approved for the treatment of AB in Germany and is used in several
other countries as an alternative therapy option for ear, nose, throat, and respiratory
tract infections [28]. Other studies employed the BSS to evaluate other treatment options in AB, e. g.
extracts of ivy leaves [35], combinations of thyme and primrose extracts [36]
[37]
[38], distillates of eucalyptus oil extract and lemon peel extract [39], and thyme/ivy combinations [40]
[41]. The generalization of the BSS as an endpoint in the different studies shows that
the score has become an established tool for consistent evaluation of AB symptoms
which could change independently from each other and across patients [23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41].
A recent example for the use of BSS – a composite scale
A recent example for the use of BSS – a composite scale
Matthys et al. [31] reported the results of a randomised, double-blind, placebo-controlled study investigating
the efficacy and tolerability of EPs 7630 tablets by means of the BSS. A total of
406 adults with AB were randomly assigned to one of four treatment groups (three different
doses of EPs 7630 and one placebo) and treated for seven days. Primary outcome measure
was the change of the BSS between baseline and day 7. As expected in a self-limiting
disease, BSS decreased in all groups but showed a significant greater reduction in
the groups treated with EPs 7630 (p < 0.0001). This result was in line with the other
patient reported outcomes, i. e. the IMOS (Integrative Medicine Outcome Scale) and
the IMPSS (Integrative Medicine Patient Satisfaction Scale). IMOS and IMPSS are 5-point
verbal rating scales, with IMOS describing the general health status of the patient
(rated independently by both the investigator and the patients) and IMPSS rated by
the patients and mirroring their satisfaction with treatment. Both scales have been
used in several clinical studies assessing treatment outcome of AB [22]
[24]
[25]
[26]
[31]
[32]
[33]
[34] and the European Committee for Homeopathy takes them into account for data collection,
too [42]. Overall, the BSS total score, the IMPSS and the IMOS are in line with patient reported
outcomes [31] as has been shown in earlier trials [28]
[33].
Validation of the BSS
Every composite endpoint has to fulfill validity criteria to be applied as a suitable
outcome measure. In case of the BSS, validity has been analysed by Lehrl resulting
in a comprehensive report [43]. The essence of this report is submitted for publication elsewhere. Here we give
a short overview of the validation of the BSS.
In total, 7 published randomized, double-blind controlled trials with EPs 7630 for
which a complete data set was on file were used to validate the BSS as an outcome
measure. For this, 2,033 patients (full analysis set) with AB were included [24]
[25]
[26]
[28]
[31]
[32]
[33]. The BSS is designed to assess the treatment success of AB. When analysing the scale
different aspects of validity were implied according to the classical test theory:
(1) content validity, (2) construct validity and (3) criterion validity, with the latter being subdivided into (3a) concurrent validity and (3b) predictive (prognostic) validity [44].
(1) For content validity of the BSS it is necessary that the symptoms included in the scale reflect the expert
knowledge about AB. A further prerequisite, especially for valid application, is that
all users of the scale, including family physicians, responsible for the treatment
of the vast majority of patients with CC/AB, are able to rate the symptoms correctly.
In conclusion, high content validity was reported for the BSS.
For (2) construct validity, generation of the scale has to be checked, i. e. it is crucial to find out if the
instrument is in agreement with its underlying theory. Looking at the BSS, the five
bronchitis-specific items must reflect the whole spectrum of the impact on health
of patients suffering from AB. To ensure this univariate nature of AB, factor analyses
(data from 7 studies [24]
[25]
[26]
[28]
[31]
[32]
[33]) were performed resulting in two groups of factors which varied independently from
each other ([Fig. 1]). The first factor, named “cough domain”, includes the four symptoms cough, rales,
chest pain and dyspnoea. The second factor only comprises the factor “sputum domain”
and is named accordingly. Both factors proved to be independent i. e. a certain score
(degree of severity) of the “cough domain” does not imply a corresponding severity
score of the “sputum domain”. Hence, the two factors can be designated as autonomous.
Fig. 1 The subscales “cough domain” and “sputum domain” (for the y-axis: courses of medians
for severity) may result in different scores during the course of the disease (data
from Chuchalin et al. [33]).
The main question for (3) criterion validity is if the scale really measures the parameters it is supposed to measure. Concerning
the BSS, it should reflect the degree of severity of AB. Several criteria are involved,
some of them concerning concurrent, other predictive validity. (3a) Concurrent validity was confirmed by means of statistical tests, i. e. the self-assessments by the patient
of the 3 symptoms “cough”, “sputum production” and “chest pain” were in line with
the investigators assessments of the same symptoms [43]. Because of the combination of objective and subjective items that constitute the
BSS, the assessment of the general practitioner can be done only in conjunction with
the subjective feedback of the patient.
For (3b) predictive validity of the BSS, two prerequisites have to be fulfilled: 1) Even without medication, the
BSS decreases in the week following the first visit. 2) Through treatment with an
effective preparation, e. g. EPs 7630 in a double-blind, randomized, placebo-controlled
trial, the BSS decreases more rapidly than through placebo treatment. According to
the available studies [24]
[25]
[26]
[28]
[31]
[32]
[33], both predictions apply for the BSS and thus, the BSS meets the requirements of
(3a) concurrent validity and (3b) predictive validity.
Finally, generalisation of the BSS can be stated, since assessment with the BSS appears
to be largely independent of any special training for the assessing doctors, of their
nationality, native language, age, or sex. This is also true for the patients being
assessed: according to the analyses of Lehrl there is no indication for any systematic
effect of patients’ nationality, age, and sex on the outcome of the BSS [43]. Thus, the BSS can be recommended as a valid measurement tool for the severity of
AB ([Table 1]) and its validity has been recently accepted by the European Medicines Agency’s
(EMA) Committee on Herbal Medicinal Products (HMPC) [45].
Table 1
Relationship between BSS score and severity of acute bronchitis [43].
|
BSS total score
|
Clinical interpretation
|
|
0
|
No acute bronchitis
|
|
1 to 2
|
Acute bronchitis unlikely
|
|
3 to 7
|
Mild acute bronchitis
|
|
8 to 12
|
Moderate acute bronchitis
|
|
13 to 17
|
Severe acute bronchitis
|
|
18 to 20
|
Very severe acute bronchitis
|
Limitations of the BSS
One major result of the validity check-up was the differentiation of the above mentioned
subscales “cough domain” and “sputum domain” [43]. As shown in [Fig. 1] and [Fig. 2], sputum production may stay the same or even increase during the first days of the
disease whereas the symptoms combined in the “cough domain” decrease constantly. However,
this is in line with clinical reality: in everyday routine patients usually report
an important relief, if after 2 – 3 days the cough becomes wet for a few days.
Furthermore, it was shown that EPs 7630 is very effective for cough relief when compared
with placebo [25]
[26]
[31]. If once sputum production starts, coughing is no more a complaint for the patient
and expectoration subsides rapidly (sputum score = 0 at day 7 ([Fig. 2]); i. e. no more phlegm will be coughed up, which is consistent with the clinical
impression).
Fig. 2 Independent trend of the two domain scores (for the y-axis: courses of medians for
severity) when all 1,955 patients of the seven studies are included (active drug:
n = 1,181; placebo: n = 774) [43].
Thus, both complexes alone appear to reflect reality even better than the BSS total
score and we recommend using these subscales to assess disease severity in AB. To
receive values similar to the ones shown in [Table 2], the four variables of the “cough domain” must be summed up and the total divided
by four, because they are interrelated. Clinical interpretation of the results according
to [Table 2] and [Fig. 2] is then possible.
Table 2
Relationship between the score of the two BSS subscales and severity of acute bronchitis
[43].
|
Score of BSS subscales
|
Clinical interpretation
|
|
0
|
No acute bronchitis
|
|
–
|
Cough
|
Sputum
|
|
> 0 and < 1.5
|
Mild
|
Mild
|
|
1.5 to 2.4
|
Moderate
|
Moderate
|
|
2.5 to 3.4
|
Severe
|
Severe
|
|
3.5 to 4.0
|
Very severe
|
Very severe
|
Both a weakness and strength of the BSS could be its subjective, patient reported
character. However, in a self-limiting disease patient related outcomes are the most
important ones and no other patient related outcomes for CC/AB are available.
Overall, we consider the BSS as appropriate for the assessment of treatment response
in AB. Validation is based on well-designed and placebo-controlled trials and the
results have been published in peer-reviewed journals. Nevertheless, we emphasize
the need for further suitable studies employing the BSS with treatment options other
than EPs 7630, which was used in most of the trials analysed here. Also, future trials
should iterate even more on relating the BSS to clinically relevant changes of disease
severity.
Beyond randomized controlled trials
Beyond randomized controlled trials
Recently Matthys and Kamin reviewed all studies employing the BSS until April 2012
[46]. The authors state that improvement in the BSS found in these studies correlates
with clinical outcomes and patients’ satisfaction with treatment. Thus, they recommend
further use of the BSS as a reliable and convenient clinical trial tool for selecting
and evaluating patients in studies of AB. But does the validated BSS help in clinical
practice, too? Both for research and clinicians the evidence based BSS helps to determine
which are clinically important symptoms in CC/AB. These ones should be assessed in
individual patients. They help if the course of the disease is unusual, protracted
and/or the decision on further diagnosis and treatment are pending [23]. Thus, the validated BSS, and especially its subscales, are recommended tools helping
the doctor to focus the consultation and better manage patients with CC/AB, whatever
is the prescribed therapy.
In conclusion, a validated BSS assessing pharmacotherapeutic effects in CC/AB is now
available. Using the subscales “cough domain” and “sputum domain” separately is considered
to be even more helpful. They are useful and important tools for future drug trials
in CC/AB, but they also help physicians to focus their consultation in many patients
with CC/AB.[2]
