Rapid Responses to:

QUALITY IMPROVEMENT REPORTS: Jennifer E Ball, Jane Osbourne, Sarah Jowett, Mike Pellen, and Mark R Welfare Quality improvement programme to achieve acceptable colonoscopy completion rates: prospective before and after study BMJ 2004; 329: 665-667 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Quality Improvement in Colonoscopy

Robert O Hart   (20 September 2004)

Improving colonoscopic performance

Peter M Irving, Joel E.D. Mawdsley, Richard J. Makins   (22 September 2004)

We need more competant colonoscopists not fewer

Michael Graham Bramble, John Silcock, Matt Rutter, Ann Powell, Ken Matthewson, John Painter   (5 October 2004)

There is more than one way to skin a cat

Mark R Welfare   (7 October 2004)

Improvement and confidence in colonoscopy completion

Alistair S McIntyre, David A Gorard   (8 October 2004)

Should the less successful endoscopists be removed or retrained?

Christiana Laban, Akil Elewa - Associate Specialist in Surgery   (2 November 2004)

Response to Quality improvement programme to achieve acceptable colonoscopy completion rates: prospective before and after study

M Hanif Shiwani, Hamed N. Khan   (20 November 2004)

Quality Improvement in Colonoscopy 20 September 2004
Robert O Hart, Consultant Surgeon, Day Surgery Lancashire Teaching Hospitals NHS Tust, Royal Preston Hospital, Preston, Lancs, PR2 9HT Send response to journal: Re: Quality Improvement in Colonoscopy	Sir, The Authors are to be congratulated on this constructive attempt to address the problem of colonoscopy completion rates that are, at a national level, woeful. However, there are several points that arise. Firstly, there are the criteria used to assess completion. Only entry to the terminal ileum is an infallible marker: A diverticulum can impersonate the appendix orifice; a loop of transverse colon can, under some circumstances, trans-illuminate in the right iliac fossa; the angulation at the hepatic flexure can be mistaken for the tri-radiate fold and an otherwise undistinguished fold of mucosa can be mistaken by the unwary for the ileo-caecal valve. Therefore completion assessed by only one of these criteria must be regarded as suspect. To qualify as complete without ileal entry at least two and preferably three of the criteria should have been demonstrated, not only to the endoscopist but to a third party such as the endoscopy nurse. Secondly, the authors laudably comment that exclusion criteria should enable “…a better measure of the technical skills of endoscopists.” They then include equipment failure. It is an old adage that a poor workman blames his tools. Since November 2001 I have performed between 200 and 400 colonoscopies per year, approximately 1000 in three years. In that series, with a 94% completion rate, only one failure has been put down to equipment failure, and I blame myself for that. I would suggest that any endoscopist whose completion figures are significantly affected by equipment failure should have the true reasons for that failure very closely examined. Then there are their appointment times. The authors claimed to book 2 colonoscopies and 6 to 8 gastroscopies to a list, with 20 minute appointments: Their arithmetic doesn’t add up to a 3.5 hour list. Presumably they just go home early – pleasant when it happens but surely not something you should plan for when the resource in question is in such demand. The authors discovered that the most successful colonoscopists were doing the least colonoscopies. I would be fascinated to discover if that finding applied anywhere else, since it is so counter-intuitive. Indeed, extrapolated to an extreme it would seem to argue that the best results would be obtained by the most inexperienced and untrained operators on their first and only attempt, which is patently ridiculous. The authors decided to “… concentrate the colonoscopies in the hands of the more successful colonoscopist”. Laudable indeed. Perhaps they could share with us the secrets of how to deal with an ageing and irascible if highly respected consultant colleague who has completion rates in the order of 30% but refuses point blank either to give it up (“It’s an essential part of my specialisation”) or retrain (“I’m too old/busy for that” or “It’s not me, there’s something wrong with your figures”). Furthermore, many endoscopy units are provided with equipment at a level no greater the minimum compatible with delivery of the service as it is. Alternating lower and upper GI endoscopy on the same list may well be the only practical way to keep the list going while equipement is sterilised. While this is not necessarily a consideration in purely colonoscopy lists, the natural consequence of concentrating colonoscopies on one list is to concentrate other endoscopies onto other lists: The supply of endoscopes in some departments may be insufficient to cope. Finally, it would be interesting indeed to see an analysis of the financial consequences of non-completion of colonoscopy. I may be cynical but I believe that our managers are more likely to be persuaded by a balance sheet than merely clinical considerations. Yours Sincerely Robert O Hart Competing interests: None declared
Improving colonoscopic performance 22 September 2004
Peter M Irving, Research Fellow Research Centre for Gastroenterology, Barts and The London School of Medicine, London E1 2AD, Joel E.D. Mawdsley, Richard J. Makins Send response to journal: Re: Improving colonoscopic performance	Sir, We read with interest Ball and colleagues’ paper on improving caecal intubation rates at colonoscopy (1) in which they conclude that these improvements were due to three key measures; increasing appointment times; allocating the procedures to the most skilled operators; and improving bowel preparation in frail patients. Despite the fact that their paper highlights important issues and describes admirable attempts to address them, we would like to raise issues related to each of their interventions. Colonoscopy appointment times were increased from 20 to 30 minutes. The Royal College of Physicians recommends that consultant gastroenterologists perform a maximum of six colonoscopies per notional half day (3 1/2 hours) (2); 30 minutes is therefore slightly less than the suggested minimum time per procedure. Perhaps increasing appointment times even further would have resulted in even better caecal intubation rates. Moreover, an appointment shorter than the recommended lower limit is unlikely to be adequate for training purposes. Secondly, although it may seem sensible to allocate colonoscopies to the most proficient practitioners, this intervention could also have an impact upon training of juniors. Thirdly, admitting frail patients for bowel preparation may not be a cost-effective measure. The authors do not state how many extra admissions this created, however, as 14% of patients attending for colonoscopy are 75 or older (3), frail patients requiring admission may represent a significant burden for many hospitals. Computerised tomography without bowel preparation is likely to identify gross pathology in such patients and may be a viable alternative. In addition, inpatient bowel preparation is, in our experience, often less effective than that performed at home: we agree with the authors that admission to wards with expertise in this area is important. Finally, there have been significant improvements in colonoscopic technology over recent years, such as carbon dioxide (instead of air) insufflation, and variable stiffness colonoscopes, which are likely to improve patient comfort and completion rates. The authors do not state whether any of their equipment was updated between audit periods. 1. Ball JE, Osbourne J, Jowett S, Pellen M, Welfare MR. Quality improvement programme to achieve acceptable colonoscopy completion rates: prospective before and after study BMJ 2004; 329: 665-667 2. Royal College of Physicians of London. Consultant Physicians Working for Patients. 2nd edition. London. Royal College of Physicians. 2001 3. Bowles CJA, Leicester R, Romaya C, Swarbrick E, Williams CB, Epstein O. A prospective study of colonoscopy practice in the UK today: are we adequately prepared for national colorectal screening tomorrow? Gut 2004;53: 277-83 Competing interests: None declared
We need more competant colonoscopists not fewer 5 October 2004
Michael Graham Bramble, Consultant Gastroenterologist James Cook University Hospital Middlesbrough TS4 3BW, John Silcock, Matt Rutter, Ann Powell, Ken Matthewson, John Painter Send response to journal: Re: We need more competant colonoscopists not fewer	Quality improvement programme to achieve acceptable colonoscopy completion rates: prospective before and after study. Br Med J 2004;329:665-7 Editor - We would like to commend Dr Welfare’s group on identifying an area of clinical practice requiring service improvement and then having the ability to improve this over two audit cycles despite the multidisciplinary nature of endoscopy. They acknowledge that they were starting from a low base and identified several reasons why this might be, including the allocation of just 20 minutes for each colonoscopy. We would argue that, considering that this is a JAG (Joint Advisory Group) registered training unit, even 30 minutes seems a little short unless training lists are identified separately. Most units allow 30 minutes for non-training lists and less than this should be seen as wholly inadequate. Of more interest is that the colonoscopists doing the most colonoscopy had a worse completion rate than those doing fewer. The reasons for this are not stated but in some units this relates to an over- reliance on trainees performing the colonoscopies. The paper also doesn’t state whether this was a separate variable from time allocated. It is important to make this distinction as those with larger lists might be trying to do the impossible in the allotted time. It is also important for other reasons, as this data flies in the face of evidence from surgical practice. Although one individual example is given of substantial improvement, the range of completion rates in the second audit cycle ranged from 34% to 100%. No data is given regarding anonymised individual performance or the effect of including trainees Clearly the improvement in completion rate is substantial but how much of this is due to the better-trained colonoscopists and how has trainee performance been factored in? A major drawback of concentrating colonoscopy in fewer hands is that these individuals will be unable to meet the increasing workload arising from the proposed colorectal cancer-screening programme. We believe a similar improvement could have been achieved by better training alone, enabling even inexperienced colonoscopists to achieve a high standard quickly. It might have been better to re-train the individuals doing less well rather than losing them from the pool of doctors able to perform such procedures. Improved training is the key to long-term improvement in colonoscopy skills and there are now three national and seven regional sites. Each centre has been running ‘basic skills in colonoscopy’ training courses for some time and the universal experience has been that a high completion rate is possible after a relatively short period of time, providing the skills required are taught and put into practice. We also feel that nurse endoscopists with the correct training will achieve the same colonoscopy success rate at gastroenterologists and colorectal surgeons. With the advent of screening for colorectal cancer we will require many more nurse colonoscopists than at present and relegating nurse endoscopists to ‘only gastroscopy’ does the profession no favours. We need to ensure that more nurses are trained to perform this examination to a high standard otherwise the proposed screening programme will undoubtedly fail. The big question is how long will it take the training centres to correct the deficiency that we know exists in many units as a consequence of the colonoscopy audit before embarking on the training of the next generation of nurse endoscopists? Yours etc Mike Bramble1 Gastroenterologist John Silcock1 Gastroenterologist Ann Powell1 Nurse Endoscopist Ken Matthewson2 Gastroenterologist Matt Rutter3 Gastroenterologist John Painter4 Gastroenterologist 1. James Cook University Hospital, Marton Road, Middlesbrough TS4 3BW 2. Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP 3. North Tees University Hospital, Hardwick, Stockton on Tees 4. Sunderland Royal Infirmary, Kyall Lane, Sunderland Faculty, North East England, Endoscopy Training Centre, James Cook University Hospital Competing interests: None declared
There is more than one way to skin a cat 7 October 2004
Mark R Welfare, Consultant/Senior lecturer North Tyneside General Hospital, NE29 8NH Send response to journal: Re: There is more than one way to skin a cat	The responses by Hart and Bramble et al contribute to the discussion on how nationally we can improve colonoscopy completion rates. Firstly, completion is better verified by terminal ileal intubation or alternatively perhaps by agreement between endoscopist and attending endoscopy nurses that the caecem has been reached. secondly, equipment failure was a very rare cause of failure and may reflect instrument abuse. Thirdly, 2 colonoscopies and 8 gastroscopies is equivalent to 12 'points' and plenty for a full list given that emergencies may need to be added to the end. We need to move more to quality in endoscopy and away from doing as many cases as is humanly possible. Similarly, endoscopy units need to be funded to the point of quality, which may need more investement to allow all-colonoscopy lists. Balance sheets may sway managers but costs of litigation are potentially high for missed diagnosis. Fourthly, it may well be that total number of colonoscopies performed in a lifetime is a guide to experience and likely to influence completion rates but many experienced clinicians have additional responsibilities and may not be working full time in clinical medicine. This may explain the apparent paradox between success completion rates and number performed per year. The lowest completion rates were not by trainees in gastroenterology who all had very respectable completion rates. Hopefully Bramble et al feel reassured that we are trying to fulfil the spirit of the guidelines for training by reinforcing 'guidance that a consultant should be available in the endoscopy department when colonoscopy was undertaken by any trainee, even those in their final year of training, as per JAG guidelines'. I agree that reducing the number of colonoscopists is potentially problematic. However, with rearrangements of lists it may be possible to get more colonoscopies done by the best operators and not diminish the total number performed. Retraining and continued training is certainly important and as mentioned in the paper, colonsocopists 'undertook to undergo further training to maintain skills. Three consultant colonoscopists have now attended the St Marks training course and another has done a course on teaching colonoscopy'. One of our consultants has taught on the very valuable courses run at national and regional level. However, retraining may not be suitable for everyone given other demands on their time for continued professional development. Bramble et al are ideally placed to gather data on the effect of training on performance. I suspect that the biggest effects of training will be on new endoscopists and that it may be difficult, but not impossible, to improve the completion rates of long-standing endoscopists with poor habits or who do not have the aptitude. I agree that nurse endoscopists may well be the way forward and one of our specialist nurses is training in colonoscopy. Nurse endoscopists were mentioned in the article not with the intention of relegating them to gastroscopy only but to highlight the fact that by rearranging lists to have colonoscopy-only lists their was a need for someone else with appropriate skills to perform the endoscopies. Sorry if it appeared that the skills of nurse endoscopists were being down-played. In conclusion, individual units need to find their own ways of dealing with low completion rates in their units. We have found a method of improving colonoscopy completion rates in our unit but we do not hold it up as the only solution. It is likely that other units will need other solutions. My belief is that the process of audit alone will probably contribute to improving completion rates by making individuals aware if they are not performing satisfactorally. Pride is a powerful motivator. Hopefully our article and the responses will stimulate that national debate. Competing interests: None declared
Improvement and confidence in colonoscopy completion 8 October 2004
Alistair S McIntyre, Consultant Gastroenterologist Wycombe Hospital, High Wycombe, Bucks HP11 2TT, David A Gorard Send response to journal: Re: Improvement and confidence in colonoscopy completion	Sir, We read with interest the encouraging report showing a quality improvement in colonoscopy completion rates which yielded similar results to our own recently published study 1. Both studies indicated that sufficient time has to be given for colonoscopists to achieve good completion rates and trying to do too many cases on a list is detrimental to quality. However, where the report by Ball et al. concentrated colonoscopies on those operators with good colonoscopy completion rates, we were able to show that colonoscopists who were not performing sufficient procedures to keep their skill levels high could improve their completion rates by undertaking more colonoscopies. Our study included data on trainees at varying levels of experience and demonstrated that with appropriate training good completion rates are possible even in this environment. Ball et al. showed an improved completion due to a reduction in failures caused by poor bowel prep from 9% to 2.6%. It would be interesting to know whether this was all the result of their change in practice with in-patients perhaps all being cared for by experienced nurses, or whether the bowel preparation improvement was also seen in out- patients. Our experience suggested that as the Unit focused on the quality of colonoscopy there were fewer failures due to poor bowel prep and we attributed this to a general enthusiasm to ensure proper preparation rather than the instigation of specific measures. Finally the data are not analysed statistically although this is possible. The ability to reach the caecum can be analysed using binomial statistics to develop 95% confidence intervals and compared with tests such as c 2. Insufficient data are given in the paper to allow accurate calculation but approximate limits can be quantified. Thus during cycle 2 (2000 – 2002) it is stated that colonoscopists with 90% completion rates were doing less than 8 colonoscopies a month. This suggests that they were at most completing colonoscopies in 86 of 96 cases per year with calculated 95% confidence intervals of 81.7 to 94.9% {ie 78 to 92 of the 96 cases}. Colonoscopists doing 21 cases per month completed less than 83 % implying 209 completed colonoscopies of 256 annual procedures giving 95% confidence intervals of 76.3 to 86.2% {ie 195 to 221 of the 256 cases}. Thus statistically there was little difference between the two {chi square test p > 0.05}. Presumably for colonoscopists with intermediate numbers and rates the overlap was even greater. The authors have shown improved departmental colonoscopy completion rates but it is far harder to show statistical improvement for individuals when relatively low numbers of procedures are being undertaken. Our view is that colonoscopy completion rates should be given with confidence intervals if results are to be compared. Reference 1 Gorard DA, McIntyre AS. Completion rate to caecum as a quality measure of colonoscopy in a District General Hospital Colorectal Disease 2004: 6: 243 – 249. Competing interests: None declared
Should the less successful endoscopists be removed or retrained? 2 November 2004
Christiana Laban, Surgical SpR Poole Hospital NHS Trust, BH15 2JB, Akil Elewa - Associate Specialist in Surgery Send response to journal: Re: Should the less successful endoscopists be removed or retrained?	EDITOR - Ball et al should be commended for highlighting their low colonoscopy completion rates, and performing an audit to identify areas for improvement. We agree that the use of audit in colonoscopy is invaluable and the points highlighted in the study were valid, with undoubted relevance to other colonoscopy units. The paper notes a considerable interoperator variation in completion rates (varying from 34% to 100%). They do not tell us whether the least successful colonoscopists were trainees or consultants, and if trainees, whether they were supervised or not during the endoscopy sessions. Surely, removing the worst performers from colonoscopy without the opportunity of further training or supervised sessions is merely ensuring they never have the chance to improve their skills. The best way to improve performance is through training. Would it not have been preferable to educate the worst performers, either by way of attending formal colonoscopy courses, or with colonoscopy sessions supervised by the best performers? Removing the worst performers from the equation is a quick fix, resulting in an automatic improvement in the statistics without solving the heart of the problem. Although the authors' actions resulted in an increase in completion rates, this is merely a short-term solution to the problem. What will happen to the completion rates we wonder when the best performers leave their current positions or retire, leaving only the poorer performers who have been starved of further training? The paper highlighted one individual whose completion rates had improved from 79% to 95% over a four-year period. It is a shame that the poorer performers were not given the chance to increase their skills likewise. Competing interests: None declared
Response to Quality improvement programme to achieve acceptable colonoscopy completion rates: prospective before and after study 20 November 2004
M Hanif Shiwani, Consultant Surgeon Barnsley General Hospital, S75 2EP, Hamed N. Khan Send response to journal: Re: Response to Quality improvement programme to achieve acceptable colonoscopy completion rates: prospective before and after study	Sir, We read with interest the recent article on improving colonoscopic completion rates at North Tyneside General Hospital (1). Two completed audit cycles appeared to improve completion rates from 60% to 88%. This success was dependent on adjusting both operator and patient factors. However, we would like to highlight some issues. The two audit cycles are over a long duration of three years with different sample size in each cycle resulting in poor comparison of like with like. Completion was variably defined as either visualising the caecum or intubating the ileum. Terminal ileal intubation requires more time, especially in the learning curve. The effect of time on successfully intubating the terminal ileum is not mentioned. Since colonoscopy is operator dependent, excluding staff with low completion rates alone may have altered the overall rates. Similarly the addition of new endoscopists with better skills may improve department’s completion rates without changing practice, the authors ignored this. Patient factors such as overall colonic size, recto-sigmoid size, presence of diverticulosis and redundancy of the transverse colon contribute to difficult colonoscopy and therefore its completion rate (2). These factors were not mentioned in the article and overcoming such factors would require experience gained over the period of the audit cycle and not specifically by improving bowel preparation as mentioned in the article. Data on the scopes used are not available as technically difficult colonoscopy can be improved by the use of small calibre variable stiffness scopes (3). Changing the scopes over time is a well-known practice in many modern endoscopic units. This may have affected the completion rate. This article fails to prove that the claimed success was specifically due to changes to practice as suggested by the authors. Success may have been due to multiple other factors independent of the changes in practice. Thank you. References: 1-Ball JE, Osbourne J, Jowett S, Pellen M, Welfare MR. Quality improvement programme to achieve acceptable colonoscopy completion rates: prospective before and after study. BMJ. 2004 Sep 18; 329(7467): 665-7. 2- Saunders BP, Halligan S, Jobling C, Fukumoto M, Moussa ME, Williams CB, Bartram CI. Can barium enema indicate when colonoscopy will be difficult? Clin Radiol. 1995 May; 50(5): 318-21. 3-Horiuchi A, Nakayama Y, Kajiyama M, Fujii H, Tanaka N. Usefulness of a small-caliber, variable-stiffness colonoscope as a backup in patients with difficult or incomplete colonoscopy. Am J Gastroenterol. 2004 Oct; 99(10): 1936-40. Authors: MR H Khan MRCS SpRegistrar General Surgery Barnsley District General Hospital Mr M H Shiwani FRCS Eng, FRCS Glasg., FRCSI [Gen.Surg] Consultant General Surgeon Barnsley District General Hospital Honorary Senior Clinical Lecturer University of Sheffield Email: mhshiwani@aol.com Competing interests: None declared