Comment We would like to thank Thackray® Ltd for permission to reproduce the illustration from their catalogue. J HOLLINGWORTH FRCS
Surgical Registrar H D KAUFMAN ChM FRCS Consultant Surgeon Selly Oak Hospital Birmingham References I Bennett-Jones MJ. A retractor for cholecystectomy. Lancet 1955;2:854. 2 Temple JG. The Bennett-Jones retractor. 7 R Coll Surg Edinb 1976;21:237-8.
Primary restorative colectomy in malignant left-sided large bowel obstruction The paper by Dorudi, Wilson and Heddle (Annals, November 1990, vol 72, p393) clearly demonstrates that primary restorative colectomy may be performed in certain cases of malignant left-sided large bowel obstruction, without the need for ontable colonic lavage. Disappointingly, the authors do not help the reader to decide which patients would benefit from this procedure. No details are given about the general condition of the patients, the amount of faecal loading of the colon or the degree of distension of the obstructed bowel. We are not told whether the surgeons encountered any difficulty when suturing bowel ends of greatly disparate circumferences or whether any bowel was of doubtful viability and how they dealt with this. I remain unconvinced that this procedure can be safely performed on all patients with malignant left-sided large bowel obstruction. There must surely be some cases where the patient is unfit and the additional operating time taken to perform a difficult anastomosis (particularly after a low anterior resection) would jeopardise the patient's life. In such instances the patient is better served by the formation of a proximal colostomy (and a distal mucous fistula wherever possible). There must also be cases where the caecum is so greatly distended and ischaemic that an extended right hemicolectomy with ileocolic anastomosis is more appropriate. The authors are to be congratulated for their enviable results: a mortality rate of less than 1%, no wound infections and no clinical anastomotic leaks; commendable results indeed for emergency surgery. It must be remembered, however, that this paper reports a very small case series and I would be most interested in the results of the next 18 patients treated. Mr Dorudi and his colleagues have raised an interesting issue and clearly large prospective randomised studies are required. Although they have successfully challenged the view that large bowel preparation is required before anastomosis, I will await further results before adopting this procedure for all patients with large bowel obstruction that I manage. HAROUN GAJRAJ MS FRCS Lecturer in Surgery St Thomas' Hospital London
Blood transfusion in total hip replacement: is it always necessary? I read with interest the paper by Porteous and Miller (Annals, January 1991, vol 73, p44) concerning the necessity for blood transfusion after total hip replacement. They are to be con-
199
gratulated for furthering the awareness that blood transfusion after major surgery should be based on need rather than habit. I would, however, take issue with your assessor's less than enthusiastic comments. A mean postoperative haemoglobin concentration of 10.3 g/dl at 48 h and 11.1 g/dl at 14 days hardly counts as significant anaemia. In cardiac surgery, where increasing efforts are being made to reduce the requirements for homologous blood transfusion, the criteria for postoperative homologous transfusion in haemodynamically stable patients are as low as a haemoglobin of 8.0 g/dl or a haematocrit of 25% (1). It is also entirely possible that some of the pharmacological agents, such as aprotinin, tranexamic acid and desmopressin acetate (DDAVP), attracting interest for reducing homologous blood transfusion in cardiac surgery, could be applied in this setting of elective major orthopaedic surgery. RUSSELL MILLNER MB BS FRCS Research Registrar in Cardiothoracic Surgery St George's Hospital London Reference I Scott WJ, Kessler R, Wernly JA. Blood conservation in cardiac surgery. Ann Thorac Surg 1990;50:843-51.
Deaths following trauma: an audit of performance We read with interest the above paper by Phair et al. (Annals, January 1991, vol 73, p53). It was recommended that "more detailed studies and comparisons with other centres ... are required". This can be best achieved by a national coordinating system. We are pleased to report that the United Kingdom Major Trauma Outcome Study was established in 1989 at the North Western Injury Research Centre (NWIRC). Over 30 hospitals nationwide participate and the database now contains information on 8000 injured patients. Injury scaling is performed at NWIRC preventing intercoder variability. Statistical feedback using the TRISS methodology (1) is provided on a regular basis and is used to highlight patients for interdisciplinary audit, hopefully leading to improvements in trauma care. Additional benefits include the potential for comparative studies between hospitals employing different systems of trauma care and the development of a large database which can be used to refine the scoring systems themselves. Two further points should be clarified. Firstly, M values for less than 100 patients are not statistically reliable and consequently the Z value cannot be viewed with any confidence. Secondly, Mr Montague in his comments advocates the use of the 1990 version of the Abbreviated Injury Scaling publication in the TRISS calculations. As there are, as yet, no published regression coefficients for use with AIS90 the 1985 version should still be used. M WOODFORD MTOS Coordinator S HOLLIS Senior Medical Statistician Hope Hospital Salford References I Boyd CR, Tolson MA, Copes WS. Evaluating trauma care: The TRISS method. J Trauma 1987;27:370-8.
McBurney's point-fact or fiction? I have read with interest the letters on the siting of the appendix and incisions for its removal (Annals, January 1991, vol 73, p65). I wonder if I might be allowed to make a few comments based on embryology, anatomy and clinical observations. Many
200
Comment
aspects of treatment of this condition have changed radically since the advent of antibiotics, but we still adhere to the old principles rather than starting from basic facts. The development of the appendix and its site is quite complex. This is why it is possible for a normal appendix to lie beneath the gallbladder or anywhere below this down to the pelvis. It can also be very mobile or even transposed. In malrotation, its position is even more bizarre. The key factor is, however, the blood-vessel development and origin. The embryological origin also determines the nature of the initial pain before parietal peritoneal involvement occurs. This pain may vary in severity and even nature, and so is easily missed if a very careful history is not taken. Not surprisingly the appendix, like a Meckel's diverticulum, has a segmental level at the same level as the umbilicus. This gives us only vague help, although the fact that the appendix is developed bilaterally allows us to differentiate the source from herpes zoster and ureteric pains which never cross the midline. In choosing our incision we must know the location of the appendix. This is revealed by the point of maximum tenderness and the incision must allow for this. A rigid choice of incision does not give access if the organ is not in its 'usual site'. Once the diagnosis is made the incision chosen should be centred on this point. Next, however, we have to consider difficulties in the actual operation, even when the diagnosis is correct. If an incision is too low or too lateral, how shall we deliver the organ? We can always pull it upwards because its blood supply comes from above and similarly we can pull it safely medially, but we cannot pull it downwards or outwards. Therefore any incision will give easier access if we are too high or too medial than if we are the reverse. We can further gain advantage if we mobilise the caecum, either on its outer or inferior aspect by dividing the peritoneal fusion line formed when the colon rotates during the embryological reversal of the developmental exomphalos. In the pre-antibiotic era this would not have been safe, but now it is. In the past, any opening of planes would have been disastrous, although inside the peritoneal cavity it was probably not as dangerous as was believed-as Lawson Tait observed a century ago. What is the best incision? First, it must give access to the appendix. This means centering it over the point of maximum tenderness, whether this be near the umbilicus or out in the flank, but with the considerations outlined above. There is no perfect incision for all cases. Despite the greatest skill in preoperative diagnosis, however, we sometimes find that we have to deal with other pathology besides the appendix. For this reason it is also important to have flexibility available in our incisions. We may need to be able to extend the incision. This should be borne in mind from the outset. Dividing muscle is not always as harmful as it might seem, but it is not good to do more than necessary. For this transverse incisions can be better, quite apart from the cosmetic and mechanical benefits they may confer. A skin-crease incision with an oblique incision in the outer part of the rectus sheath is easily extended upwards and outwards, and similarly a transverse incision across the linea semilunaris can be enlarged with safety to give wide access. The paramedian and midline incisions give much less flexibility, and McBurney's presents danger as the nerves are readily cut. This classical incision is always difficult if wide vision is needed, as to give a wide peritoneal hole the funnel shape requires a larger skin incision than most other incisions. It also does not lend itself to the other manoeuvres mentioned above. Many of the criticisms of incisions are due to poor surgery. Thus, Battle's incision was often found wanting because several nerves were divided. Used carefully it offers excellent access,
especially in children. McBurney's caused weakness for the same reason and because surgeons often tore the muscles badly and weakened them. These and others have been criticised for allowing spread into pristine muscle planes. Apart from the protection of antibiotics which has altered this, we should remember that in the past the cause of this infection was often the drains used. I used in pre-antibiotic times to drain the appendix stump, the pelvis and the wound! None of these should now be necessary. One still, however, sees surgeons failing to protect the abdominal wall layers from contamination whilst the peritoneum is open. The only time now when a drain is needed is (a) when there is an abscess cavity with rigid walls and (b) when there is an oedematous caecum. In both cases the drain should be one which produces a track for pus, etc, to discharge. The modern silicone tubes are, contrary to belief, quite irritant and are only equalled in this by the old soft rubber drains. It is the track which matters - this is why, fortunately, so many varied drains work almost equally well. In case (b) it is sometimes not possible to close the appendix stump, let alone bury it (a practice I personally like) and then one can obtain security by taking a cross-stitch with soft, thick, plain catgut across the open end and then turning the caecum and sewing it against the side wall of the pelvis and anchoring it with a few catgut sutures. This is much easier than trying to resect the bowel. Another often neglected way to remove the appendix is to divide the base first and close the stump. The artery is then tied allowing the appendix to be dissected out with the finger in safety. In rare cases the mucosa is found to lie free but intact. This can be removed alone with good result leaving only a small hole to close. Many of the practices in appendicitis have been adopted and equally often condemned for the wrong reasons. Thus, Fowler's position was believed to help as the pelvic peritoneum absorbed less toxins. In reality it probably was good because it allowed better excursion of the diaphragm and by giving better lung function reduced chest complications. It was later accused of causing deep vein thrombosis which was probably due to other causes such as pelvic infection. Theory does not always accord with reality and we must not make facts fit the theory-as Richard Feynman, the mathematician, said, 'Nature cannot be fooled'. There is no doubt that the greatest advance in treating appendicitis has been the advent of antibiotics. I have removed just under 3000 acute appendices and lost four. Two of these were in children before antibiotics, both treated with enemata for their pain, one in a child when penicillin had arrived (a case presenting with a subphrenic abscess) and the last in an adult with an acute perforation with general peritonitis for which only penicillin was available. It is unwise to be dogmatic about the ideal antibiotic. Overall I have found streptomycin intraperitoneally excellent. It must be put in just before closing the peritoneum as it is accompanied by a sudden exudate of serous fluid. More recently I have supplemented this with rectal metronidazole started in theatre. With this regimen it is possible to close wounds happily with a subcuticular suture and so obtain a good cosmetic result - not a feature considered important in the past. I think this sums up an important point. We have now reached the stage visualised by Moynihan years ago when he said surgery was safe for the patient. However, it does not excuse us from now applying ourselves to a new aspect of our treatment-the finesse of our methods. G T WATTS FRCS 4 Amesbury Road Moseley, Birmingham
Surgical Registrar H D KAUFMAN ChM FRCS Consultant Surgeon Selly Oak Hospital Birmingham References I Bennett-Jones MJ. A retractor for cholecystectomy. Lancet 1955;2:854. 2 Temple JG. The Bennett-Jones retractor. 7 R Coll Surg Edinb 1976;21:237-8.
Primary restorative colectomy in malignant left-sided large bowel obstruction The paper by Dorudi, Wilson and Heddle (Annals, November 1990, vol 72, p393) clearly demonstrates that primary restorative colectomy may be performed in certain cases of malignant left-sided large bowel obstruction, without the need for ontable colonic lavage. Disappointingly, the authors do not help the reader to decide which patients would benefit from this procedure. No details are given about the general condition of the patients, the amount of faecal loading of the colon or the degree of distension of the obstructed bowel. We are not told whether the surgeons encountered any difficulty when suturing bowel ends of greatly disparate circumferences or whether any bowel was of doubtful viability and how they dealt with this. I remain unconvinced that this procedure can be safely performed on all patients with malignant left-sided large bowel obstruction. There must surely be some cases where the patient is unfit and the additional operating time taken to perform a difficult anastomosis (particularly after a low anterior resection) would jeopardise the patient's life. In such instances the patient is better served by the formation of a proximal colostomy (and a distal mucous fistula wherever possible). There must also be cases where the caecum is so greatly distended and ischaemic that an extended right hemicolectomy with ileocolic anastomosis is more appropriate. The authors are to be congratulated for their enviable results: a mortality rate of less than 1%, no wound infections and no clinical anastomotic leaks; commendable results indeed for emergency surgery. It must be remembered, however, that this paper reports a very small case series and I would be most interested in the results of the next 18 patients treated. Mr Dorudi and his colleagues have raised an interesting issue and clearly large prospective randomised studies are required. Although they have successfully challenged the view that large bowel preparation is required before anastomosis, I will await further results before adopting this procedure for all patients with large bowel obstruction that I manage. HAROUN GAJRAJ MS FRCS Lecturer in Surgery St Thomas' Hospital London
Blood transfusion in total hip replacement: is it always necessary? I read with interest the paper by Porteous and Miller (Annals, January 1991, vol 73, p44) concerning the necessity for blood transfusion after total hip replacement. They are to be con-
199
gratulated for furthering the awareness that blood transfusion after major surgery should be based on need rather than habit. I would, however, take issue with your assessor's less than enthusiastic comments. A mean postoperative haemoglobin concentration of 10.3 g/dl at 48 h and 11.1 g/dl at 14 days hardly counts as significant anaemia. In cardiac surgery, where increasing efforts are being made to reduce the requirements for homologous blood transfusion, the criteria for postoperative homologous transfusion in haemodynamically stable patients are as low as a haemoglobin of 8.0 g/dl or a haematocrit of 25% (1). It is also entirely possible that some of the pharmacological agents, such as aprotinin, tranexamic acid and desmopressin acetate (DDAVP), attracting interest for reducing homologous blood transfusion in cardiac surgery, could be applied in this setting of elective major orthopaedic surgery. RUSSELL MILLNER MB BS FRCS Research Registrar in Cardiothoracic Surgery St George's Hospital London Reference I Scott WJ, Kessler R, Wernly JA. Blood conservation in cardiac surgery. Ann Thorac Surg 1990;50:843-51.
Deaths following trauma: an audit of performance We read with interest the above paper by Phair et al. (Annals, January 1991, vol 73, p53). It was recommended that "more detailed studies and comparisons with other centres ... are required". This can be best achieved by a national coordinating system. We are pleased to report that the United Kingdom Major Trauma Outcome Study was established in 1989 at the North Western Injury Research Centre (NWIRC). Over 30 hospitals nationwide participate and the database now contains information on 8000 injured patients. Injury scaling is performed at NWIRC preventing intercoder variability. Statistical feedback using the TRISS methodology (1) is provided on a regular basis and is used to highlight patients for interdisciplinary audit, hopefully leading to improvements in trauma care. Additional benefits include the potential for comparative studies between hospitals employing different systems of trauma care and the development of a large database which can be used to refine the scoring systems themselves. Two further points should be clarified. Firstly, M values for less than 100 patients are not statistically reliable and consequently the Z value cannot be viewed with any confidence. Secondly, Mr Montague in his comments advocates the use of the 1990 version of the Abbreviated Injury Scaling publication in the TRISS calculations. As there are, as yet, no published regression coefficients for use with AIS90 the 1985 version should still be used. M WOODFORD MTOS Coordinator S HOLLIS Senior Medical Statistician Hope Hospital Salford References I Boyd CR, Tolson MA, Copes WS. Evaluating trauma care: The TRISS method. J Trauma 1987;27:370-8.
McBurney's point-fact or fiction? I have read with interest the letters on the siting of the appendix and incisions for its removal (Annals, January 1991, vol 73, p65). I wonder if I might be allowed to make a few comments based on embryology, anatomy and clinical observations. Many
200
Comment
aspects of treatment of this condition have changed radically since the advent of antibiotics, but we still adhere to the old principles rather than starting from basic facts. The development of the appendix and its site is quite complex. This is why it is possible for a normal appendix to lie beneath the gallbladder or anywhere below this down to the pelvis. It can also be very mobile or even transposed. In malrotation, its position is even more bizarre. The key factor is, however, the blood-vessel development and origin. The embryological origin also determines the nature of the initial pain before parietal peritoneal involvement occurs. This pain may vary in severity and even nature, and so is easily missed if a very careful history is not taken. Not surprisingly the appendix, like a Meckel's diverticulum, has a segmental level at the same level as the umbilicus. This gives us only vague help, although the fact that the appendix is developed bilaterally allows us to differentiate the source from herpes zoster and ureteric pains which never cross the midline. In choosing our incision we must know the location of the appendix. This is revealed by the point of maximum tenderness and the incision must allow for this. A rigid choice of incision does not give access if the organ is not in its 'usual site'. Once the diagnosis is made the incision chosen should be centred on this point. Next, however, we have to consider difficulties in the actual operation, even when the diagnosis is correct. If an incision is too low or too lateral, how shall we deliver the organ? We can always pull it upwards because its blood supply comes from above and similarly we can pull it safely medially, but we cannot pull it downwards or outwards. Therefore any incision will give easier access if we are too high or too medial than if we are the reverse. We can further gain advantage if we mobilise the caecum, either on its outer or inferior aspect by dividing the peritoneal fusion line formed when the colon rotates during the embryological reversal of the developmental exomphalos. In the pre-antibiotic era this would not have been safe, but now it is. In the past, any opening of planes would have been disastrous, although inside the peritoneal cavity it was probably not as dangerous as was believed-as Lawson Tait observed a century ago. What is the best incision? First, it must give access to the appendix. This means centering it over the point of maximum tenderness, whether this be near the umbilicus or out in the flank, but with the considerations outlined above. There is no perfect incision for all cases. Despite the greatest skill in preoperative diagnosis, however, we sometimes find that we have to deal with other pathology besides the appendix. For this reason it is also important to have flexibility available in our incisions. We may need to be able to extend the incision. This should be borne in mind from the outset. Dividing muscle is not always as harmful as it might seem, but it is not good to do more than necessary. For this transverse incisions can be better, quite apart from the cosmetic and mechanical benefits they may confer. A skin-crease incision with an oblique incision in the outer part of the rectus sheath is easily extended upwards and outwards, and similarly a transverse incision across the linea semilunaris can be enlarged with safety to give wide access. The paramedian and midline incisions give much less flexibility, and McBurney's presents danger as the nerves are readily cut. This classical incision is always difficult if wide vision is needed, as to give a wide peritoneal hole the funnel shape requires a larger skin incision than most other incisions. It also does not lend itself to the other manoeuvres mentioned above. Many of the criticisms of incisions are due to poor surgery. Thus, Battle's incision was often found wanting because several nerves were divided. Used carefully it offers excellent access,
especially in children. McBurney's caused weakness for the same reason and because surgeons often tore the muscles badly and weakened them. These and others have been criticised for allowing spread into pristine muscle planes. Apart from the protection of antibiotics which has altered this, we should remember that in the past the cause of this infection was often the drains used. I used in pre-antibiotic times to drain the appendix stump, the pelvis and the wound! None of these should now be necessary. One still, however, sees surgeons failing to protect the abdominal wall layers from contamination whilst the peritoneum is open. The only time now when a drain is needed is (a) when there is an abscess cavity with rigid walls and (b) when there is an oedematous caecum. In both cases the drain should be one which produces a track for pus, etc, to discharge. The modern silicone tubes are, contrary to belief, quite irritant and are only equalled in this by the old soft rubber drains. It is the track which matters - this is why, fortunately, so many varied drains work almost equally well. In case (b) it is sometimes not possible to close the appendix stump, let alone bury it (a practice I personally like) and then one can obtain security by taking a cross-stitch with soft, thick, plain catgut across the open end and then turning the caecum and sewing it against the side wall of the pelvis and anchoring it with a few catgut sutures. This is much easier than trying to resect the bowel. Another often neglected way to remove the appendix is to divide the base first and close the stump. The artery is then tied allowing the appendix to be dissected out with the finger in safety. In rare cases the mucosa is found to lie free but intact. This can be removed alone with good result leaving only a small hole to close. Many of the practices in appendicitis have been adopted and equally often condemned for the wrong reasons. Thus, Fowler's position was believed to help as the pelvic peritoneum absorbed less toxins. In reality it probably was good because it allowed better excursion of the diaphragm and by giving better lung function reduced chest complications. It was later accused of causing deep vein thrombosis which was probably due to other causes such as pelvic infection. Theory does not always accord with reality and we must not make facts fit the theory-as Richard Feynman, the mathematician, said, 'Nature cannot be fooled'. There is no doubt that the greatest advance in treating appendicitis has been the advent of antibiotics. I have removed just under 3000 acute appendices and lost four. Two of these were in children before antibiotics, both treated with enemata for their pain, one in a child when penicillin had arrived (a case presenting with a subphrenic abscess) and the last in an adult with an acute perforation with general peritonitis for which only penicillin was available. It is unwise to be dogmatic about the ideal antibiotic. Overall I have found streptomycin intraperitoneally excellent. It must be put in just before closing the peritoneum as it is accompanied by a sudden exudate of serous fluid. More recently I have supplemented this with rectal metronidazole started in theatre. With this regimen it is possible to close wounds happily with a subcuticular suture and so obtain a good cosmetic result - not a feature considered important in the past. I think this sums up an important point. We have now reached the stage visualised by Moynihan years ago when he said surgery was safe for the patient. However, it does not excuse us from now applying ourselves to a new aspect of our treatment-the finesse of our methods. G T WATTS FRCS 4 Amesbury Road Moseley, Birmingham
