Article Text

Download PDFPDF

Management and assessment of cough

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


1H. Fathi, 2P. W. Dettmar, 2V. Strugala, 2H. Dettmar, 1C. Wright, 1A. H. Morice. 1Department of Academic Medicine, University of Hull, Castle Hill Hospital, Cottingham, UK, 2Technostics Ltd, The Deep Business Centre, Hull, UK

Background: Cough is a very common symptom prompting a medical opinion. In the absence of other obvious chest disease, chronic cough is a diagnostic challenge. Online clinics aim to help by providing targeted information in response to specific answers to an online questionnaire. We aimed to make the European Respiratory Society guidelines on the management of cough more practical, while at the same time providing information about the condition for both patients and physicians at the distance of a click!

Methods: On a secure provider, patients were required to complete a questionnaire if over 18 years of age. A normal chest x ray within a year was necessary and history of haemoptysis was an exclusion criterion. An algorithm was applied after assigning weighting factors to the patients symptoms, to give a percentage of the maximum probability of having the three main causes of chronic cough: reflux, asthma or rhinitis. Based on guidelines, the best approach was suggested to the patient in a letter to their GP. The patient was asked to complete a follow-up questionnaire at 2 months to assess the current situation and response.

Results: Up until July 2008, 14 037 patients (58% women, mean age of 45 years) registered with the site. 8884 patients completed the questionnaire. 46.48% were diagnosed with reflux, 38.31% and 15.21% with asthma and rhinitis, respectively. 1066 (12%) completed the follow-up questionnaire out of whom 94% found the site easy to use, 68% took the recommended medication, 73% found it helpful and 60% suggested it helped in having a better communication with their GP.

Conclusion: A web-based approach to the guidelines makes them more practical and handy. The virtual world of the internet is going to have a major impact on health services and web clinics could save time and money.


1P. A. Marsden, 1U. Alam, 2A. A. Woodcock, 2J. A. Smith. 1University Hospital of South Manchester NHS Foundation Trust, Manchester, UK, 2University of Manchester, Manchester, UK

Rationale: Cough is the most common condition for which patients seek medical advice. Prospectively, over 6 months we aimed to investigate adherence to current BTS guidelines in patients referred to a specialist cough clinic.

Methods: Data were collected prospectively on all new referrals to a tertiary level cough clinic, from January to June 2007. Information was gathered regarding source of referral, age, gender, time coughing, provisional diagnosis at the time of referral, imaging, lung function, empirical treatment trials and the presence of angiotensin-converting enzyme (ACE) inhibitors.

Results: 52 new patients were seen (71% female); mean age 57.5 years (±12.4); median 27.5 months coughing (range 3–240); 67% were referred from secondary care. At the point of referral, 77% had a chest x ray, 46% an HRCT thorax and 67% spirometry, but 69% still had no provisional diagnosis. The majority (92%) of patients had undergone at least one treatment trial (76.5% of primary and 100% of secondary care referrals). Completed trials of treatment included proton pump inhibitor (52%), bronchodilator (17%), oral corticosteroids (19%), inhaled corticosteroids (67%), nasal corticosteroids (76.9%). In addition, 8% were taking an ACE inhibitor (12% of primary and 6% of secondary referrals).

Conclusions: The application of BTS guidelines for the investigation of cough in primary and secondary care seems patchy, especially the recommendations on chest x ray and spirometry. The development and implementation of local treatment pathways and referral guidelines may improve patient outcomes.


A. Hamilton, P. Marsden, S. Decalmer, H. Sumner, A. Kelsall, K. McGuinness, A. Woodcock, J. A. Smith. University of Manchester, Manchester, UK

Objectives: Interstitial lung disease (ILD) patients frequently complain of cough but little is known about objective cough frequency in this group of conditions. We compared objectively measured cough in ILD patients to the subjects own assessment of their cough and against objective data for healthy volunteers and chronic cough patients.

Methods: 10 ILD patients (median age 70 (range 63–83), two female, TLC% 65% (49–96), DLCO% 35.5% (25–54) and smoking history of 28 pack-years (0–52)), were selected at random from a specialist clinic at the North-West Lung Centre, Manchester. Using the VitaloJAK cough monitor, we collected 24-h cough recordings for each subject, which were manually counted (explosive cough sounds per hour). Participants also completed day and night visual analogue scales and the Leicester cough questionnaire. Comparative objective data for healthy volunteers (n  =  9, median age 59 years, seven female) and chronic cough patients attending a specialist clinic (n  =  86, median age 58 years, 59 female) were obtained from previous studies.

Results: Median cough rate in ILD was 11.2 coughs/h (range 1.8–36.7), with daytime median 14.1 coughs/h (1.9–47.9) and overnight median 2.7 coughs/h (range 0–19.7). Strong correlations were found between objective cough rates and visual analogue scales for day (r  =  0.83, p = 0.003) and night (r  =  0.84, p = 0.002) and between 24-h cough rates and the Leicester cough questionnaire (r  =  −0.71, p = 0.02). Cough rates were not significantly correlated with disease severity (as assessed by pulmonary function); however, they were correlated with time since diagnosis (r  =  0.89, p = 0.001). Cough rates in this patient group were significantly higher than those of healthy volunteers (median 0.67 coughs/h, range 0.16–2.89, p<0.001) but similar to rates seen in subjects presenting with isolated chronic cough (median 15.7 coughs/h, range 0.3–140.8, p = 0.22).

Conclusion: In ILD patients, cough frequency is comparable to that seen in subjects presenting with chronic cough. VAS scores and cough-related quality of life closely correlated with objective cough frequency, suggesting subjects were highly aware of this symptom, perhaps due to the unpleasantness of coughing in the context of worsening breathlessness. However, cough frequency did not appear to relate to disease severity but did worsen with disease duration.


S. Faruqi, V. Mann, W. Sheedy, R. Thompson, C. Wright, A. H. Morice. Department of Academic Medicine, University of Hull, Castle Hill Hospital, Cottingham, UK

Background: Cough is the commonest symptom for which patients seek medical advice yet the study of cough has been hampered by a lack of clinically useful, well-validated measurement tools. The assessment of chronic cough has been improved by the development of objective ambulatory cough monitoring systems and subjective quality of life questionnaires. Experimental induction of cough is a useful tool in the assessment of the cough reflex and capsaicin is a commonly used agent for the same. We wanted to assess the correlation between these measurements and their reproducibility.

Methods: This was a prospective observational study in 25 patients with chronic idiopathic cough of greater than 6 months duration. All patients had an initial 24-h cough recording. They also completed a Leicester cough questionnaire (LCQ), a validated cough-related quality of life questionnaire and had a capsaicin cough challenge performed. They were reviewed at 8 weeks when all three assessments were repeated. The cough recorder comprised a Sony Hi-MD Walkman and cardioid microphone (secured to the patient’s chest) and a 1 GB Hi-MD minidisc. Raw data were converted to WAVE format and analysed using Cubase SX and Audacity software.

Results: The study included 25 patients (15 women) with a mean age of 54 years. All three forms of assessment (the 24-h cough recording, LCQ score and capsaicin sensitivity) were found to be highly reproducible at 8 weeks (p<0.01). The findings are summarised in the table. There was good correlation seen between the three parameters (p<0.05, ρ = 0.5–0.9). Good negative correlation was observed between median C2 and C5 and the total number of coughs in both the visits (p<0.05).

Conclusion: Objective cough counting, subjective assessment of cough-related quality of life and capsaicin cough reflex sensitivity are reproducible forms of assessment of chronic cough. Subjective and objective assessment of cough are highly correlated. Sensitivity to capsaicin correlates with increased numbers of objectively quantified coughs. Total coughs represent mean number of coughs recorded in the 24-h period. The LCQ scores vary from 3 to 21, with higher scores representing better quality of life. C2 and C5 are the concentrations of nebulised capsaicin inhaled that induce ⩾2 and ⩾5 coughs, respectively. The median C2 and C5 are depicted.

Abstract P64 Table


1S. S. Biring, 2N. Y. Yousaf, 2S. Matos, 2I. D. Pavord. 1Kings Hospital, London, UK, 2Glenfield Hospital, Leicester, UK

The normal range of cough frequency is unknown. We undertook a study to measure 24-h ambulatory cough frequency using the Leicester cough monitor in adult volunteers with normal spirometry and no reported respiratory symptoms.

52 adults (32 female) were recruited ranging from 20 to 80 years old, with a mean (SD) age of 47 years (15). They all underwent 24-h ambulatory cough frequency monitoring followed by assessment of cough sensitivity using incremental inhaled concentrations of capsaicin. The concentrations at which the volunteer coughed two times (C2) and five times (C5) were recorded. Cough frequency was measured as the number of individual coughs over 24 h.

Mean age for women was 44 years (12) and men 50 years (19). The geometric mean (log SD) number of cough episodes in 24 h was 22.4 (0.51). Women coughed more than men (30.9 (0.39) vs 14.8 (0.6); 2.1-fold; 95% CI 1.04 to 4.17; p = 0.036). Cough frequency was similar in the six current smokers and was not related to body mass index or age. There was no correlation between 24-h cough frequency and C2 (r  =  −0.08) or C5 (r  =  −0.03).

In conclusion, cough frequency in normal volunteers is higher in women than men. It is not obviously different in healthy current smokers, with ageing or in obese volunteers. There is no correlation with capsaicin cough sensitivity, suggesting that these measurements are assessing independent aspects of the cough phenotype.


1T. F. P. McKeagney, 2L. Polley, 3R. Costello, 2J. MacMahon, 1L. P. McGarvey. 1Royal Victoria Hospital , Belfast, UK, 2Belfast City Hospital, Belfast, UK, 3Beaumont Hospital, Dublin, Republic of Ireland

Introduction and Objectives: Cough reflex hypersensitisation is known to be a key clinical feature in patients with chronic troublesome cough. Bouts of coughing are often triggered by low threshold stimuli encountered during normal daily activities1 and this self-reported sensory hyperresponsiveness has been shown to correlate with objective measures of cough reflex sensitivity.2 This is perceived by patients to be a disruptive aspect of their condition and leads to impaired quality of life.3 We set out to try and identify the prevalence and potential clinical features associated with this sensory hyperresponsiveness.

Methods: We undertook a retrospective case note review of 200 sequential referrals to a specialist cough clinic between 2001 and 2004. Self-reported sensory hyperresponsiveness was defined as cough aggravated by: 1 change in air temperature; 2 aerosols, scents or deodorants; 3 laughing, talking or singing.

Results: 135 charts (67.5%) had the required information recorded. Of these patients 72% (n  =  97) were women. Only 6.5% were current smokers. The majority (85%, n  =  115) had a clear aetiology for their cough and the remaining 15% were deemed to have idiopathic cough. Sensory hyperresponsiveness was present in 63% of patients (n  =  85) and was more frequently reported by women (N  =  70, p = 0.001). There was no significant difference in age, cough duration, atopic status or preceding viral upper respiratory tract infection between the two groups.

Conclusions: Self-reported sensory hyperresponsiveness was present in the majority of patients attending this specialist cough clinic and was more prevalent in women.


1V. Mann, 1S. Faruqi, 2R. Wilmot, 1A. H. Morice. 1Department of Academic Medicine, University of Hull, Castle Hill Hospital, Cottingham, UK, 2Department of Biochemistry, Hull Royal Infirmary, Hull, UK

Background: Lactate dehydrogenase (LDH) is an enzyme found in almost all tissues of the body and serum levels of this enzyme can be elevated in several conditions. This enzyme is found in abundance in the heart, liver, red blood cells, kidneys, muscle, brain and lungs. Five different LDH isozymes are known (LDH1–LDH5). We had observed raised serum LDH levels in a subgroup of patients with chronic cough, which we investigated further.

Methods: Patients seen at our clinic with chronic idiopathic cough prospectively had a serum LDH assay done. Both total and specific isozyme levels were checked. Patients with other co-morbidities that could possibly lead to raised LDH levels were excluded.

Results: Forty-one patients (30 women, mean age 61 years) were included in the study. Out of these 33% had their LDH values above the normal laboratory reference range of 215–485 μ/l. The mean LDH was 461 (range 311–678, SD 79). Among those in whom the total LDH was within the normal reference range, the majority were towards the upper limit of normal. 68% of the patients had LDH values in the fourth quartile of the reference range or above. This increase in LDH was predominantly due to a rise in isozymes 4 and 5. The mean LDH4 and LDH5 percentages were 14% and 20%, respectively. In the subgroup with raised total LDH the mean LDH4 and LDH5 percentages were similar at 13% and 22%, respectively. Normal ranges for LDH4 are from 6% to 12% and those for LDH5 from 3% to 17%. This pattern of rise in LDH suggests a muscle origin for the enzyme.

Conclusion: Serum LDH levels are elevated in a substantial proportion of patients with chronic cough. This rise seems to be due to a release of muscle LDH. This could possibly occur due to the muscular strain involved in the act of persistent coughing. These findings warrant further investigation.


1S. Decalmer, 2A. Kelsall, 2K. McGuinness, 1A. Woodcock, 1J. A. Smith. 1Manchester University, Manchester, UK, 2North West Lung Research Centre, Manchester, UK

Rationale: In determining the cause of chronic cough, patient-reported responses to treatment trials are often used, but this has never been validated by objective cough counts.

Aim: To compare patient perceived responses with changes in cough frequency and quality of life (Leicester cough questionnaire; LCQ).

Methods: In a prospective study, chronic cough patients (n  =  100) were investigated for asthma (PFT ± methacholine challenge and sputum eosinophils), post nasal drip (ENT examination) and oesophageal reflux (impedance) and treatment trials were then targeted to the relevant triggers (8 weeks). Cough monitoring (VitaloJAK, manual counting) and the LCQ were completed prior to investigation and posttreatment. Patients’ global rating of response was categorised as: good response—disappearance/substantial reduction in cough; partial response—some improvement in cough; no response—cough at/above pretreatment levels

Results: All patients completed pretreatment measures. A global rating of response was available in 94 subjects with 82 completing LCQ scores and 43 cough recordings posttreatment. A good global response was reported by 46 (49%) patients, partial in 23 (24%) and no response in 25 (27%). There were no significant differences in age, gender or pretreatment measurements between these groups (see table). Statistically significant improvements in LCQ were seen in the good and partial responders (good > partial responders (p = 0.004) and partial > no response (p = 0.01)). Cough rates only improved significantly in good responders. Cough rate improvements were greater in good than partial responders (p = 0.01) but there was no difference between partial and non-responders (p = 0.91). Improvement in cough rate correlated with improvement in LCQ (r  =  −0.68, p<0.001).

Conclusions: Half of patients with investigation-based treatment trials reported a good response, with substantial improvements in cough rate and LCQ, the magnitude of which were far greater than in patients who reported a partial or no response. Partial responders reported significant LCQ improvements compared with non-responders, despite a similar small reduction in cough rate; potential explanations are improvement of cough intensity rather than frequency; improvement in psychological state with thorough investigation and reassurance; hypervigilance to small improvements in cough frequency and a placebo effect influencing the LCQ scores.

Abstract P68 Table
View Abstract