Eur Respir J 2009; 34: 819–824

DOI: 10.1183/09031936.00126908
CopyrightßERS Journals Ltd 2009

The online Cough Clinic: developing

guideline-based diagnosis and advice
P.W. Dettmar*, V. Strugala*, H. Fathi#, H.J. Dettmar*, C. Wright# and A.H. Morice#

ABSTRACT: The aim of the present study was make chronic cough guidelines more practical and AFFILIATIONS
*Technostics Ltd, The Deep Business
user friendly by developing an internet-based interactive diagnostic questionnaire for chronic
Centre, Hull, and
cough. #
Cardiovascular and Respiratory
A prospective cohort study of chronic cough sufferers was conducted in the UK, following Studies, Hull York Medical School,
European Respiratory Society guidelines for the diagnosis and management of chronic cough. Castle Hill Hospital, Cottingham, UK.
Depending on the response to 16 specific questions, the medical condition responsible for the
CORRESPONDENCE
patient’s chronic cough was ascertained according to a predetermined diagnostic algorithm A.H. Morice
designed to differentiate the three common causes of chronic cough. Appropriate advice and Hull York Medical School
treatment recommendations were then provided. University of Hull
8,546 adults with chronic cough completed the Cough Clinic diagnostic questionnaire. 46.1% Castle Hill Hospital
Castle Road
were suggested to have reflux, 38.7% asthma and 15.2% rhinitis. Participants found the website Cottingham
easy to use (94%), the advice helpful (73%) and that it helped them to communicate with their HU16 5JQ
general practitioner better (60%), and 62% reported taking the recommended treatment. UK
E-mail: [email protected]
The Cough Clinic, an internet-based diagnostic site for chronic cough, had a large uptake by
chronic cough sufferers in the UK. Almost half were diagnosed as having reflux as the probable Received:
cause of their chronic cough. Internet diagnosis by expert algorithm provides a novel mechanism Aug 18 2008
for patients to access guideline-recommended therapies and enhances dialogue between Accepted after revision:
patients and physicians. March 17 2009
First published online:
March 26 2009
KEYWORDS: Chronic cough, diagnosis, gastro-oesophageal reflux, questionnaire

he challenge of making an accurate diag- medical care is sought [6]. In the absence of any

T nosis remains one of the most fulfilling roles

of physicians [1]. Clinical practice guidelines
are considered as important tools to guide
obvious respiratory disease, three common causes
have been proposed for chronic cough; reflux
disease, an asthma syndrome and rhinitis. The
evidence-based decision making [2]. However, the European Respiratory Society (ERS) guidelines
routine implementation of guidelines, many of illustrate the development of the approaches to
which can be of hundreds of pages, has been cough [7, 8].
reported to be low [3]. Doctors and patients are The Cough Clinic (www.coughclinic.org.uk;
becoming more proficient in using the internet, National Health Service/University of Hull,
with 31.8 million adults in the UK estimated to Hull, UK) is an internet-based diagnostic site
have internet access (15.2 million households), with developed to suggest a probable diagnosis of the
searching for health information at the top of the condition causing a patient’s chronic cough. The
list of reasons for using the web [4]. Google1 Cough Clinic individualises advice based upon
searchers can find the correct diagnosis for their the guidelines set by the ERS [7] thus making the
medical condition in 58% of cases [5]. Hence, the guidelines more practical, while at the same time
hypothesis is that a guideline-based web clinic providing information about the condition for
would be a practical way to address patients’ both patients and their physicians.
healthcare queries.
The diagnosis of the aetiology of cough in a patient
Chronic cough is defined as a cough lasting for requires relevant information from the patient’s
.8 weeks and is a disabling symptom with a medical history followed by the assigning of
significant impact on a patient’s quality of life. weighting factors to the information to predict
Cough is the commonest symptom for which the most likely diagnosis. Correct diagnosis is
European Respiratory Journal

This article has supplementary material accessible from www.erj.ersjournals.com

Print ISSN 0903-1936
Online ISSN 1399-3003 c
EUROPEAN RESPIRATORY JOURNAL VOLUME 34 NUMBER 4 819
COUGH P.W. DETTMAR ET AL.

crucial for an effective therapy. An online internet-based system Certain symptoms (e.g. haemoptysis) were identified as ‘‘red
that claims to be able to give an accurate diagnosis should be flag’’ symptoms and the completion of the questionnaire was
able to perform all of the above functions. terminated with advice to obtain specialist help. Red flag
symptoms were only seen in eight patients (three coughing up
Here we describe the Cough Clinic website and its role in blood and five coughing up more than one cupful of phlegm).
diagnosis of chronic cough and patient appreciation of the
pathway. Algorithm and diagnosis
An algorithm was applied to the questionnaire responses, after
METHODS assigning weighting factors, to calculate a score as a percentage
Data collection of the maximum probability of the three main causes of
On entering the Cough Clinic website the patient was chronic cough. The condition with the highest percentage was
registered, provided consent and demographic data was chosen as the most probable diagnosis (reflux, asthma or
rhinitis) causing the chronic cough. The algorithm was scored
collected (age, sex and region of residence). Before proceeding
as a percentage of the total possible score for each symptom.
with the questionnaire, the patient was required to confirm
Thus, although there were more questions pertaining to reflux
that a normal chest radiograph had been obtained, as it is
the algorithm was not biased towards a reflux diagnosis since
mandatory for the investigation of chronic cough.
each individual question contributed a smaller percentage to
Ethical approval for the study was granted by the local the total.
research ethics committee for Hull and East Riding, UK.
Research and development approval was granted by Hull Treatment
National Health Service Trust, Hull, UK. All patients provided When a diagnosis is made, the ERS guidelines recommend a
online consent at the time of completion of the Cough Clinic therapeutic trial of the most appropriate treatment. The Cough
questionnaire. Clinic generated a generic letter for the patient to take to their
primary care physician that referred to the completed ques-
The questionnaire was set over three pages. The patient was tionnaire, the ERS guidelines and the suggested diagnosis of
asked about the length of time that they have suffered with the cause of the patient’s chronic cough; it also suggested the
cough, and a Likert scale (0–10) used to score the severity of the most suitable treatment trial for the patient (see online
cough. The patient was asked to answer a series of questions to supplementary material).
determine other relevant factors. The second page gathered
further information on the patient’s smoking habits and medica- Follow-up
tion use. The third page asked about key diagnostic symptoms As for any other medical consultation, the clinician needs to
and their severity by Likert scale (0–5) (table 1). For the diagnosis assess the effectiveness of the diagnosis and the recommended
of reflux-induced cough, these questions were based on the treatment. After 2 months the patient received a request to
validated reflux symptom index (RSI) [9] and the Hull reflux complete a follow-up questionnaire over the internet. This
cough questionnaire (HRCQ) [10]. Asthma symptoms were allowed for the assessment of the patient’s current symptoms,
based on Royal College of Physicians questionnaire [11]. current medications, patient compliance and the efficacy of the

TABLE 1 Questions used to ascertain the probable medical condition responsible for a patient’s chronic cough

Number Question Weighting

1 Cough with eating (during or straight after meals) Reflux

2 Cough with certain foods Reflux
3 Cough when you get out of bed in the morning Reflux
4 Cough brought on by singing or speaking (for example, on the telephone) Reflux
5 Hoarseness or a problem with your voice Reflux
6 Clearing your throat Reflux
7 Cough after lying down Reflux
8 Heartburn, chest pain, indigestion or stomach acid coming up Reflux
9 Wheezing or chest tightness in general Asthma
10 Cough waking you from sleep Asthma
11 Shortness of breath when not coughing Asthma
12 Blocked or stuffy nose Rhinitis
13 Excess mucus in the throat, or dripping down the back of the nose Rhinitis
14 Itchy nose and/or sneezing Rhinitis
15 Loss of the sense of smell Rhinitis
16 A feeling that mucus is running down the back of your throat Rhinitis

The severity of the above key diagnostic symptoms was rated on a Likert scale (0–5). The questions were asked in a random order and not as listed.

820 VOLUME 34 NUMBER 4 EUROPEAN RESPIRATORY JOURNAL

P.W. DETTMAR ET AL. COUGH

recommended treatment. We also inquired into the patients’ cough (table 1). Each question was scored for severity using a
satisfaction with the Cough Clinic. Likert scale of 0–5.
Depending on the score given for each question, a weighting
RESULTS
was applied relevant to the importance of that symptom for the
Demographics
three common causes of chronic cough.
Between the time the site went live (January 2006) and October
2007, 13,610 people had registered at the site, of which 8,546 The breakdown of any symptoms reported (score 1–5) by the
(63%) patients in the UK completed the Cough Clinic online population group studied is shown in figure 2; it was possible
cough diagnosis questionnaire. There was a female dominance to identify differences between the groups. There was a clear
(57%) and the mean¡SD age was 45.5¡16.4 yrs, with ages distinction in the type of symptoms reported by those
ranging from 18–86 yrs (fig. 1). diagnosed with probable asthma and those diagnosed with
probable rhinitis. Focusing on the moderate to severe
The duration of cough, for all of the responders, was
symptoms (score 3–5), which had the weighting factors applied
.3 months; in 64.8% of patients it was ,1 yr and 13.7% of
to them, shows that each predicted diagnosis presents with a
patients had had their cough for .5 yrs.
different profile (fig. 3). Those with probable rhinitis
Patients were asked to score the severity of their cough on a responded highly to questions 12–16, those with probable
Likert scale of 0–10 (10 being the most severe and frequent asthma responded to questions 9–11, whereas those with
cough) and the mean¡SD score for cough among the patients probable reflux were those that did not respond high to
was 5.85¡2.1. questions 9–16 but had a high proportion of symptoms related
to questions 1–8.
The mean¡SD cough score varied depending on their
suggested diagnosis (p,0.0001, one-way ANOVA), with those Coughing after lying down was reported to be the most severe
patients diagnosed with reflux scoring 5.4¡2.2, those diag- accompanying symptom (20% scored it as 5 and 18.5% as 4),
nosed with rhinitis scoring 5.8¡2.08, and those diagnosed while hoarseness was the mildest accompanying symptom
with asthma scoring 6.4¡2.15. (15.8% scored it as 1 and 17.4% as 2). Clearing the throat in
probable reflux patients, coughing after lying down in
41.4% of patients were either current smokers or had probable asthmatic patients, and post-nasal drip in probable
previously been smokers. 11% of responders had previous rhinitis were the most troublesome symptoms in the different
diagnosis of respiratory illnesses, such as chronic obstructive groups.
pulmonary disease (COPD), bronchiectasis, emphysema and
asthma, with asthma the greatest (9% of all). Validation study
The Cough Clinic algorithm was tested in a pilot study of 30
From the drug history data, 21.6% of patients were on inhalers,
patients attending the Hull cough clinic (Hull, UK). Patients
of which 11% were steroid inhalers. 19% took blood pressure
were administered the questionnaire before being seen as new
tablets, with 7% of these on angiotensin converting enzyme
patients by the medical team. There was a close association
(ACE) inhibitors. The known side-effects of ACE inhibitors
between the web-based Cough Clinic diagnosis and that of the
include chronic cough, which occurs at the rate of ,15%,
clinician’s full work up. Only two patients did not have
depending on the population studied [12].
agreement in diagnosis between the two methods. We
recognise that there is insufficient power to provide an
Specific questions to aid diagnosis
accurate estimate of diagnostic accuracy, but took the view
There were 16 questions that were used to specifically ascertain
that there would be variation between different clinicians
the medical condition responsible for the patient’s chronic
similar, if not greater than, that between the algorithm.
2000 Final diagnosis
Three probable diagnoses, reflux, asthma and rhinitis, were the
main outcomes of the questionnaire; 3,936 (95% CI 3,846–4,027)
1500 (46.1%) of patients were diagnosed with reflux, 3,310 (95% CI
3,222–3,398) (38.7%) were diagnosed with asthma and 1,300 (95%
Subjects n

CI 1,173–1,427) (15.2%) were diagnosed as having rhinitis (fig. 4).

1000
Follow-up questionaire
A request to complete a follow-up questionnaire 2 months
after the initial diagnosis was sent to 8,434 patients. Response
500
was only given by 1,047 (12.4%). Of the patients completing the
follow-up questionnaire, 94% said they found the site easy to
use, 73% found the advice helpful, while 60% stated that it
0 helped them communicate with their general practitioner (GP)
18–20 20–30 30–40 40–50 50–60 >60
better. 62% of patients had taken a recommended treatment.
Age yrs
DISCUSSION
FIGURE 1. Breakdown of ages of the population completing the question-
naire. There was no specific dominant age group.
Guidelines and approved evidence-based clinical algorithms
are vast swathes of information that are hard to memorise and c
EUROPEAN RESPIRATORY JOURNAL VOLUME 34 NUMBER 4 821
COUGH P.W. DETTMAR ET AL.

100
90
Frequency of symptom %

80
70
60
50
40
30
20
10
0
Eating

Certain foods

Get up

Speaking

Hoarseness

Throat clearing

Lying down

Heartburn

Wheeze

Sleep disturbance

Short breath

Blocked nose

PND

Sneeze

Loss smell

Mucus throat
FIGURE 2. Proportion of patients reporting the presence of symptoms from questions 1–16 (score 1–5). h: reflux; &: asthma; &: rhinitis. PND: post-nasal drip.

put into practice. The use of online tools to put guidelines into Our previous experience indicates that in chronic cough patients
practice and provide quality information to patients is a seeking a consultation the overwhelming majority have seen at
valuable addition to the healthcare armoury. The use of the least one and often two doctors [13].
Cough Clinic, an internet-based diagnostic website to diagnose
the condition causing a patient’s chronic cough, was assessed There has been large utilisation of the Cough Clinic site by adult
in the current study. chronic cough sufferers in the UK covering a wide demographic,
with the majority suffering from cough for 3 months to 1 yr. The
It should be stressed that definite diagnosis cannot be obtained 16-part symptom questionnaire and associated algorithm was the
using this internet-based approach without full investigative main diagnostic device to determine if reflux, rhinitis or asthma
work-up. The suggested diagnosis is provided with appropriate was the probable cause of the chronic cough. The questions were
advice and treatment but further medical consultation is derived from studies characterising the symptom complex
advocated, with a computer-generated letter provided. How- associated with each condition. In reflux, our previously
ever, it is more likely to propose a diagnosis that is unlikely to determined symptom profile of patients with pH-proven acid
have been put forward by the GP in an initial consultation. reflux and chronic cough was utilised [14]. For asthma, the Royal

80
Frequency of moderate to severe

70

60

50
symptom %

40

30

20

10

0
Eating

Certain foods

Get up

Speaking

Hoarseness

Throat clearing

Lying down

Heartburn

Wheeze

Sleep disturbance

Short breath

Blocked nose

PND

Sneeze

Loss smell

Mucus throat

FIGURE 3. Proportion of patients reporting the presence of symptoms from questions 1–16 (score 3–5). h: reflux; &: asthma; &: rhinitis. PND: post-nasal drip.

822 VOLUME 34 NUMBER 4 EUROPEAN RESPIRATORY JOURNAL

P.W. DETTMAR ET AL. COUGH

general population (internet users seeking medical advice

online). In addition, it is our intention to follow-up the patients
who have received an initial diagnosis for their chronic cough
Rhinitis with a second internet-based questionnaire. This follow-up
15.2%
questionnaire will be used to obtain information on treatment
efficacy and will enable us to carry out internet-based clinical
trials as well as epidemiological and market research studies.
Reflux However, to date, follow-up response has been low (12.4%);
46.1% therefore, data gathered thus far has not been presented here.
Asthma Although the current study was based only on chronic cough
38.7%
sufferers in the UK (with internet access) it is our intention to
extend the internet-based questionnaire to other countries
where guidelines are available. This will make it possible to
better study the aetiology and the epidemiology of chronic
cough across countries and populations.

STATEMENT OF INTEREST
FIGURE 4. Probable diagnosis of medical condition responsible for chronic
Statements of interest for P.W. Dettmar, V. Strugala, H.J. Dettmar and
cough in 8,546 patients completing the Cough Clinic questionnaire. A.H. Morice can be found at www.erj.ersjournals.com/misc/statements.
dtl
College of Physicians questionnaire was used [11]. For rhinitis,
questions from the rhinosinusitis-specific Sino-nasal Assessment
Questionnaire was used [15]. It should be noted that, although REFERENCES
the three questionnaire components of the Cough Clinic are 1 Pauker SG, Gorry GA, Kassirer JP, et al. Towards the simulation of
validated, the whole questionnaire and algorithm is not. There clinical cognition. Taking a present illness by computer. Am J Med
were three distinct patterns of response to the 16 questions. A 1976; 60: 981–996.
high score to questions 9–11 pointed to asthma, high scores in 2 Field MJ, Lohr KN. Clinical Practice Guidelines: Directions for a
questions 12–16 indicated rhinitis, while consistently moderate to New Programme. Washington, National Academy Press, 1990.
high scores to questions 1–8 signified reflux. 3 Woolf SH, Grol R, Hutchinson A, et al. Clinical guidelines:
potential benefits, limitations, and harms of clinical guidelines.
The population of chronic cough sufferers who completed the BMJ 1999; 318: 527–530.
Cough Clinic process is the largest cohort studied to date by an 4 Powell J, Clarke A. The WWW of the World Wide Web: Who,
order of magnitude. Previously published epidemiological What, and Why? J Med Internet Res 2002; 4: e4.
studies of chronic cough have used cohorts ranging from 43– 5 Tang H, Ng JH. Googling for a diagnosis – use of Google as a
diagnostic aid: internet based study. BMJ 2006; 333: 1143–1145.
228 patients, owing to the specialist nature of the condition and
6 Morice AH. Epidemiology of cough. Pulmon Pharmacol Therap
that the prevalence data of the three main causes of chronic
2002; 15: 253–259.
cough varies widely [16–28]. The prevalence of asthma ranged 7 Morice AH, Fontana GA, Belvisi MG, et al. ERS guidelines on the
6–59%, gastro-oesophageal reflux disease (GORD) 0–41% and assessment of cough. Eur Respir J 2007; 29: 1256–1276.
rhinitis 8–56%. We amalgamated the data from these 13 8 Morice AH, Fontana GA, Sovijarvi AR, et al. The diagnosis and
studies, and the overall number of patients was 1,258; 25% management of chronic cough. Eur Respir J 2004; 24: 481–492.
with asthma, 20% with GORD and 34% with rhinitis [8]. 9 Belafsky PC, Postma GN, Koufman JA. Validity and reliability of
the reflux symptom index (RSI). J Voice 2002; 16: 274–277.
In the study presented here the prevalence of probable GORD 10 Faruqi S, Brook H, Hunter V, et al. Reproducibility and sensitivity
was 46.1%, higher than previously observed, indicating that of the Hull Reflux Cough Questionnaire. Am J Respir Crit Care Med
those presenting to secondary care may be the ‘‘tip of the 2009; 179: A575.
iceberg’’ and that the condition may be more common than 11 Juniper EF, O’Byrne PM, Guyatt GH, et al. Development and
previously believed. Probable asthma was responsible for validation of a questionnaire to measure asthma control. Eur Respir
chronic cough in 38.7% of our study group population, which J 1999; 14: 902–907.
is in the range previously seen. Probable rhinitis was less 12 Ravid D, Lishner M, Lang R, et al. Angiotensin-concerting enzyme
prevalent in chronic cough sufferers than previously found inhibitors and cough: a prospective evaluation in hypertension
and in congestive heart failure. J Clin Pharmacol 1994; 34: 1116–
(15.2%). Our population is significantly different from those
1120.
previously reported in the literature in that we have not
13 Everett CF, Kastelik JA, Thompson RH, et al. Chronic persistent
excluded patients with significant airflow limitation. It is cough in the community: a questionnaire survey. Cough 2007; 3: 5.
possible that other diagnoses, such as COPD, may be 14 Everett CF, Morice AH. Clinical history in gastroesophageal
responsible for cough in a proportion of the patients reported cough. Respir Med 2007; 101: 345–348.
here; however, it is also possible that patients with airflow 15 Fahmy FF, McCombe A. General versus specific sino-nasal
obstruction may have one of the underlying causes of cough outcome measures. Clin Otolaryngol 2002; 27: 283–284.
discussed previously. This is one of the unavoidable limita- 16 Brightling CE, Ward R, Goh KL, et al. Eosinophilic bronchitis is an
tions of an internet-based algorithm. important cause of chronic cough. Am J Respir Crit Care Med 1999;
160: 406–410.
The study has given us an interesting insight into the likely
aetiology of chronic cough in this biased segment of the
17 Hoffstein V. Persistent cough in nonsmoker. Can Respir J 1994; 1:
40–47. c
EUROPEAN RESPIRATORY JOURNAL VOLUME 34 NUMBER 4 823
COUGH P.W. DETTMAR ET AL.

18 Irwin RS, Corrao WM, Pratter MR. Chronic persistent cough in 23 O’Connell F, Thomas VE, Pride NB, et al. Capsaicin cough
the adult: the spectrum and frequency of causes and successful sensitivity decreases with successful treatment of chronic cough.
outcome of specific therapy. Am Rev Respir Dis 1981; 123: Am J Respir Crit Care Med 1994; 150: 374–380.
413–417. 24 Palombini BC, Villanova CA, Araujo E, et al. A pathogenic triad in
19 Irwin RS, Curley FJ, French CL. Chronic cough. The spectrum and chronic cough: asthma, postnasal drip syndrome, and gastro-
frequency of causes, key components of the diagnostic evaluation, esophageal reflux disease. Chest 1999; 116: 279–284.
and outcome of specific therapy. Am Rev Respir Dis 1990; 141: 25 Poe RH, Harder RV, Israel RH, et al. Chronic persistent cough.
640–647. Experience in diagnosis and outcome using an anatomic diag-
20 Marchesani F, Cecarini L, Pela R, et al. Causes of chronic persistent nostic protocol. Chest 1989; 95: 723–728.
cough in adult patients: the results of a systematic management 26 Poe RH, Israel RH, Utell MJ, et al. Chronic cough: bronchoscopy or
protocol. Monaldi Arch Chest Dis 1998; 53: 510–514. pulmonary function testing? Am Rev Respir Dis 1982; 126: 160–162.
21 McGarvey LP, Heaney LG, Lawson JT, et al. Evaluation and 27 Simpson G. Investigation and management of persistent dry
outcome of patients with chronic non-productive cough using a cough. Thorax 1999; 54: 469–470.
comprehensive diagnostic protocol. Thorax 1998; 53: 738–743. 28 Smyrnios NA, Irwin RS, Curley FJ. Chronic cough with a history
22 Mello CJ, Irwin RS, Curley FJ. Predictive values of the character, of excessive sputum production. The spectrum and frequency of
timing, and complications of chronic cough in diagnosing its causes, key components of the diagnostic evaluation, and outcome
cause. Arch Intern Med 1996; 156: 997–1003. of specific therapy. Chest 1995; 108: 991–997.

824 VOLUME 34 NUMBER 4 EUROPEAN RESPIRATORY JOURNAL