Keyworth Health Centre clinical whooping cough monitoring study 1977 to present day
Doug Jenkinson DM FRCGP

Brief description
Keyworth Health Centre serves a discrete community 5 miles south of Nottingham, England. This medical practice under the leadership of general practitioner Dr Doug Jenkinson (myself) has recorded all detectable cases of whooping cough (based on a minimum of 3 weeks of paroxysmal coughing) since 1977. The figures provide a unique set of data about this important disease, that compliments official statistics and provides an alternative perspective.

Keyworth versus England and Wales incidence graph   
year by year number and rates Keyworth and England & Wales table
age groups affected in each year table
individual cases table

Tips for making a clinical diagnosis

The study in more detail.
The general medical practice at Keyworth Health Centre in Nottinghamshire, England, is the only practice covering Keyworth, Tollerton and some surrounding villages. The practice population has been roughly 11,000 since 1970. In 1977 there was an outbreak of whooping cough which occurred following a massive drop in the national immunisation rate. It was possible to calculate from this small outbreak affecting 191 people (126 under 5), the vaccine protection rate in the under fives, as efficacy had been called into question. The conclusion was that the vaccine appeared as effective as had been thought (about 80%) and the results were published in the British Medical Journal.

Because of the relative ease of detecting clinically diagnoseable whooping cough in a small discrete practice with enthusiastic staff, I continued recording as much detail as I could of all cases I subsequently could find. Whenever practicable, pernasal swabs were taken for pertussis, and the positive culture rate confirmed we were diagnosing correctly. Since 2002 we have been able to do a blood test for pertussis toxin IgG, and the results of these continue to support the validity of our clinical diagnosis.

I have written several papers as a result of this study, which now numbers over 700 cases, four of these have been in the BMJ. I have watched the number of cases remain steady for 25 years, while the number of notifications has dropped to under 600 per year. I became very concerned about the lack of awareness of whooping cough and its features among doctors and nurses. So five years ago I set up a website (the parent of this one this one, www.whoopingcough.net ) to inform patients about whooping cough and to to help them diagnose it for themselves. The feedback has confirmed its value and it gets a thousand visitors a day. The bonus for site visitor is that they can listen to the sound of the characteristic coughing on sound files.

I had been mulling over the possibility of another paper to compare the incidence in Keyworth with the rest of the country to demonstrate that the reasons for the low notification rate are different from those believed. Because of current discussion (July 2006) about the incidence of covert whooping cough and the danger that once again, misinformation from the media could damage vaccine uptake, I have decided to put the data from the Keyworth study on the web as it should be of interest to some health care workers and epidemiologists.

The year by year table shows the number of cases in Keyworth per year, and the rate per 100,000 population. Alongside are notifications for England and Wales and the rate per 100,000 population. The ratio of the two rates is shown. There is a table of ages affected in each year, classified the same way as the HPA website table of notifications.

Clinical diagnosis made on the basis of at least 3 weeks of paroxysmal coughing. This is unchanged since the study began. If a positive swab or positive blood test were found and there was less than 3 weeks of paroxysmal coughing, it would be excluded from the study total.

The cases table (may take a little time to load with a slow system) shows the age, sex and year of infection of all cases recorded since 1977 (696), and whether there was whooping, vomiting and cyanosis/apnoea. It also gives swab status if taken, and latterly, pertussis toxin IgG positivity if tested for, and the case unique index number. There is more explanation given on the page, but it is important to understand why there are some paradoxes. There are 10 more index numbers than cases. The difference mainly represents patients who were given index numbers in too much haste and later withdrawn after deciding it was not whooping cough. The diagnosis is very largely retrospective and it is quite common to revise a clinical diagnosis at a later stage.

The Keyworth versus England and Wales incidence graph (both equated to cases per 100,000 population) tells the story best. Notifications have dropped off but the incidence in Keyworth has remained high and steady for 25 years, although within this pattern there are still some very interesting changes such as age incidence and periodicity.

I hope this item attracts questions and comments. If so, I would like to put the interesting ones on a separate page. I would prefer to attribute correspondents, so please state your prefered designation or if you wish to remain anonymous. I can be emailed, or I can be contacted via Keyworth Health Centre +44 (0)115 9373527.

The original papers I have published on whooping cough are: