BMJ No 7109 Volume 315
Education and debate Saturday 13 September 1997
Detecting the effects of thromboprophylaxis: the case of the rogue reviews
Mark Petticrew, Susan C Kennedy
"Watson, Dr Watson!" I looked up at the muffled sound of Sherlock Holmes's voice as he stood at the window, gesticulating with his Stradivarius. Two hours of relentless arpeggios had finally ceased, and I gratefully removed the plugs of cotton wool from my ears. "A distinguished visitor is about to request admittance," he observed. "Would you kindly ask Mrs Cochrane to show him up?"
A few moments later a tall, bewhiskered gentleman, with an enormous portmanteau and a general air of exasperation, entered Holmes's study. He introduced himself as Professor Legge, an orthopaedic surgeon.
"Mr Holmes, only you can end this madness!" he moaned, sinking into the nearest armchair. Holmes's hawk-like eye ranged over his visitor, and I knew that the great detective was about to presage the discussion with a display of his deductive skills. "Well, Professor Legge," he began authoritatively, "I trust your search for systematic reviews on Medline this afternoon was productive?" Legge looked startled.
"Good Lord, Mr Holmes, how could you possibly know that I have just spent hours . . . searching for . . . ?" He began to swab his face with an extravagantly large handkerchief.
"Simple, my dear Legge. The light coating of dust on your face would indicate that you have attracted an electrical charge, caused, I suspect, through hours of vigilant study at a computer monitor. That, and your rather glassy stare." Legge nodded silently, while I enquired: "But how on earth did you deduce the reason for his search?"
 "This is likely to be a three-pipe problem" |  |
"Are you unaware that the use of meta-analyses in the pursuit of
effective health care is well established in your profession, Dr
Watson?" he retorted. I was abashed. Holmes continued: "Well, I too
am an exponent of evidence based methods. Naturally I am aware of their
value in medical practice." Professor Legge grunted sceptically and
began rummaging about in his bulging portmanteau.
"Well, Mr Holmes," he said, thumping a large volume of paper on to the occasional table. "I'd think these blasted meta-analyses were valuable too if they agreed among themselves." He leaned forward, almost triumphant. "But, you see, they don't. I've been trying to get some straight answers on effective surgical thromboprophylaxis all afternoon, and frankly, given the conflicting information in this lot," he jabbed his finger at the innocent pile of paper, "I'd be forgiven for mistrusting everything except what I've seen with my own eyes." |
"Ah, yes," Holmes reflected, "thromboprophylaxis in general and orthopaedic surgery is a rather vexed issue at present, is it not?"
I was not surprised to hear Holmes speak knowledgeably on thromboprophylaxis. I knew that the months following his apparent death at the hands of the evil Professor Moriarty were spent in chemical research in a French laboratory. I suspected now that this included work on pharmacological thromboprophylaxis. Moreover, his medical interests were wide enough for him to be claimed, in later years, by neurologists,(1) anaesthetists,(2) dermatologists,(3) and ophthalmologists(4) as one of their own. I did not doubt that those interests extended to surgery, though I confess I was surprised at his grasp of information technology, which had not, as yet, been invented.
Summary points |
"Perhaps, you would oblige us with the background to the case Professor Legge?" Holmes suggested. The professor nodded and leaned back into his armchair.
Professor Legge's problem
"As you may know, patients undergoing major surgery are at risk of thromboembolic disease. Half of orthopaedic patients receiving no prophylaxis develop deep vein thrombosis,(5) and almost a quarter of deaths after orthopaedic surgery have been attributed to pulmonary embolism.(6) "
"But surely most of these cases of deep vein thrombosis are clinically unimportant?" I interrupted.
"It is true that most cases are subclinical and resolve spontaneously, Dr Watson, but there is still significant morbidity associated with the condition," Legge rejoined. "Besides, cold hearted as it sounds, the treatment of problems associated with deep vein thrombosis costs almost half a billion guineas annually.(7) Imagine, moreover, the patient who is harmed, sometimes fatally, by thrombosis as a result of an operation, the purpose of which is to cure."
"Remind us, Legge, how the problem can be prevented," Holmes requested, growing more intrigued.
"Broadly, we have either the pharmaceutical methods (such as heparin, low molecular weight heparin, aspirin, and warfarin) or the mechanical methods (such as elasticated stockings). Prevention seems to represent the best management strategy, and yet effective thromboprophylaxis does not appear to be used routinely in high risk surgery."(6)
Holmes regarded Legge thoughtfully. "I take it that this is not simply due to some oversight on the part of British surgeons?"
"Certainly not! There are two principal reasons. Some surgeons
believe that while prophylaxis is e "And the other reason?"
"Safety, Mr Holmes. Surgeons are concerned about the risks of major
bleeding and haematoma associated with pharmacological prophylaxis. If
you've had a patient develop a major bleed during an operation, it's
not something you forget. Now that's evidence." Holmes looked up
sharply.
"But systematic reviews were developed to resolve just this kind of
uncertainty. Are there many in this field?"
"Nearly two dozen,"(5, 9-30) answered Legge
ruefully. "And this, Mr Holmes, is where my faith in meta-analysis
wavers. Disparity at all points of the compass! I'll give you an
example: this one shows that in total hip replacement dextran, heparin,
low molecular weight heparin, elasticated compression stockings, and
warfarin all reduce the incidence of deep vein thrombosis, while
aspirin, heparin, low molecular weight heparin, elasticated compression
stockings, and warfarin all prevent pulmonary embolism.(21)
And here's another showing that low molecular weight heparin is
best."(20)
"Well, that's clear enough," I suggested blithely, "all
methods work better than nothing at all." Legge gave me a rather
withering look before returning to his notes.
"And here's one examining fatal pulmonary embolism after hip
replacement, showing that none of these methods works better than no
prophylaxis.(29) And there's another reason for my mistrust
of meta-analyses." He leaned forward conspiratorially and continued
sotto voce. "Some of my colleagues detect a sinister power behind
many of the trials and reviews." Holmes sat bolt upright, his tufted
eyebrows knitted together.
"Moriarty!" he breathed in a chilling voice. Legge looked at Holmes
in surprise.
"Er . . . no. I was actually referring to the drug companies. Take
low molecular weight heparin: some of my colleagues suspect that
certain trials advocating its use were financed by the drug companies,
in a dastardly attempt to make us use expensive drugs."
Holmes became thoughtful. "Well, are the reviews any clearer with
regard to mechanical methods?" Legge again referred to his notes.
"Well, Holmes, stockings do seem to prevent deep vein thrombosis in
general surgery(31) and total hip
replacement.(20) They may also prevent pulmonary embolism,
but most trials have been too small to be sure. Many surgeons already
use elasticated stockings, but they're not popular with
patients - especially the chaps."
Legge by now was pacing anxiously. "You know, Holmes, I believe these
meta-analyses are simply dredging the depths of scientific inaccuracy.
Where will it all end, I ask myself? Meta-meta-analyses? Blast 'em
all." Just then he noticed the formidable Mrs Cochrane, who had
entered the library with a silver platter and was regarding him
severely. "I'm sorry, ma'am, for my intemperate language," he
mumbled.
"It's not the tone of your repudiation that worries me,
Professor Legge," snapped Mrs Cochrane as she rattled the tea tray,
"but its generalisability. 'Ere's your
tea."
"Professor Legge," I ventured, seeing the heat of his passion
pass, "do clinical guidelines provide any illumination on this most
trying matter?" Legge's reply was weary. "Guidelines! Don't
start me on the blessed guidelines! One lot cites as evidence a review
using indirect comparisons.(32) Another is based on a
mixture of conventional reviews, systematic reviews, and odd
t I attempted to offer some bluff words of comfort. "Still, Legge,
there's one consolation: no matter what you're doing, there's a
meta-analysis to support it. Your practice is 100% evidence based!"
A deathly silence pervaded the room. Holmes, Legge, and Mrs Cochrane
stared coldly at me.
Holmes rose from his armchair and withdrew his pipe from the pocket of
his dressing gown. "I shall now consider the case." He extracted an
ounce of shag from his worn carpet slipper. "This is likely to be a
three-pipe problem." He left the room with Legge's papers, and an
hour passed before he returned.
"I recall that a similar problem manifested itself in the case
of the Naval Treaty,"(37) he began. "I
suggested then that the principal difficulty lay in there being too
much evidence: the vital evidence was obscured by erroneous
information. Likewise with these meta-analyses."
"But why might their conclusions differ?" I pressed him.
"Consider their methodologies, Watson. One analysis might have
included a different set of trials because they had employed different
inclusion criteria, or simply a different search strategy." He picked
up a handful of the professor's papers. "Furthermore, a closer
examination shows that inappropriate methods of pooling data are
sometimes used. For example, these
reviews"(14, 21, 29, 32) - he shuffled the offending
papers - "have pooled data from similar treatment arms of trials and
thereby compared the incidence of deep vein thrombosis under different
forms of prophylaxis. The problem with that approach is that there may
be differences between the trial populations. A similarly flawed method
involves comparing the arms of different trials - say, the warfarin arm
of one trial and the stockings arm of another trial - and using this as
evidence that one method is more effective than another. Again, direct
comparisons never took place, and the trials may differ in many
respects."
"Such as the patients' characteristics, any other interventions
which they may have received, and even the quality of the study,"
suggested Legge.
"Quite so, Professor Legge," agreed Holmes. "My original
proposition was that systematic reviews reduce uncertainty. In this
case they have increased it."
"Perhaps, then Holmes, you would venture a summary?" I suggested.
"So what do you say to surgeons like myself who wish to use safe and
effective prophylaxis?" pressed Legge.
"Simply this: mechanical methods. They are safe, and they have been
shown to be effective in preventing deep vein thrombosis in patients at
moderate and high risk.(20, 30) They may even prevent
pulmonary embolism. In short, gentlemen, a judicious use of mechanical
methods and a suspicious mind regarding meta-analysis are the key to
this whole affair. You've heard, of course, of the case of the
Misleading Meta-analysis.(38) That, at least, should teach
us caution."
"Do you imply, Holmes, that we should never trust another systematic
review?" I interrupted in surprise.
"Far from it, Watson. Despite the recent plethora of antagonistic
correspondence in learned journals, a methodologically sound systematic
review remains the gold standard for the assessment of
effectiveness." He gestured to the pile of papers on the table. "On
the basis of what I have read here, there are four main indicators of a
sound review: firstly, a comprehensive literature search; secondly,
explicit, detailed, inclusion and exclusion criteria; thirdly, a
detailed assessment of the quality of the included studies; and,
fourthly, appropriate methods of pooling the data. The `Sign of
Four,' if you like, gentlemen!" He turned to me. "Is that succinct
enough for your memoirs, Watson?" I nodded. "In fact it's . . . er
. . . elementary!"
NHS Centre for Reviews and
Mark
Petticrew,
12 Park Crescent,
York YO3 7NU
Correspondence to: Dr Petticrew
mp25@york.ac.uk
References
1 Cherington M. Sherlock Holmes: neurologist?
Neurology 1987;37:824-5.
2 Maltby JR. Sherlock Holmes and anaesthesia. Can J
Anaesth 1988;35(1):58-62.
3 Dirckx JH. Medicine and literature: Sherlock Holmes and the
art of dermatologic diagnosis. J Dermatol Surg Oncol
1979;5:191-6.
4 Craig PC. Memoirs of Sherlock Holmes through the eyes of an
ophthalmologist. Trans PA Acad Ophthalmol Otolaryngol
1972;25(1):42-3.
5 Antiplatelet Trialists Collaboration. Collaborative overview
of randomised trials of antiplatelet therapy. III: Reduction in venous
thrombosis and pulmonary embolism by antiplatelet prophylaxis among
surgical and medical patients. BMJ 1994;308:235-46.
6 Campling EA, Devlin HB, Hoile RW, Lunn JN. Report of
the national confidential enquiry into perioperative deaths
1992/1993. London: NCEPOD, 1995.
7 Lowry JC. Thromboembolic disease and thromboprophylaxis in
oral and maxillofacial surgery: experience and practice. Br J
Oral Maxillofac Surg 1995;33:101-6.
8 Jones DR. Audit of attitudes to and use of postoperative
thromboembolic prophylaxis in a regional health authority. Ann R
Coll Surg Engl 1991;73:219-21.
9 Gent M, Roberts RS. A meta-analysis of the studies of
dihydroergotamine plus heparin in the prophylaxis of deep vein
thrombosis. Chest 1986;89(suppl 5):396-400S.
10 Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in
fatal pulmonary embolism and venous thrombosis by perioperative
administration of subcutaneous heparin: overview of results of
randomized trials in general, orthopaedic and urologic surgery.
N Engl J Med 1988;318:1162-73.
11 Daures JP, Schved JF, Momas I, Gril JC, Azoulay P, Gremy F.
Meta-analysis on randomised trials comparing the results of low
molecular weight heparins to those of fractioned heparins in the
prevention of deep venous thrombosis [in French]. Rev
Epidemiol Sante Publique 1989;37:363-9.
12 Gallus AS, Cade JF, Mills KW, Murphy W. Apparent lack of
synergism between heparin and dihydroergotamine in prevention of deep
vein thrombosis after elective hip replacement: a randomised
double-blind trial reported in conjunction with an overview of previous
results. Thromb Haemost 1992;68:238-44.
13 Lassen MR, Borris LC, Christiansen HM, Schott P, Olsen AD,
Sorensen JV, et al. Clinical trials with low molecular weight heparins
in the prevention of postoperative thromboembolic complications: a
meta-analysis. Semin Thromb Hemost 1991;17(suppl
3):284-90.
14 Greyer HG. Thromboprophylaxis with heparin and low molecular
weight heparin in elective hip surgery: current status and
perspectives. Semin Thromb Hemost 1991;17:336-42.
15 Nurmohamed MT, Rosendaal FR, Buller HR, Dekker E, Hommes DW,
Vandenbroucke JP, et al. Low-molecular-weight-heparin versus standard
heparin in general and orthopaedic surgery: a meta-analysis.
Lancet 1992;340:152-6.
16 Leizorovicz A, Haugh MC, Chapuis F-R, Samama MM, Boissel JP.
Low molecular weight heparin in prevention of perioperative thrombosis.
BMJ 1992;305:913-20.
17 Simonneau G, Leizorovicz A. Comparison of different methods of
prevention of venous thrombosis in orthopaedics. Ann Fr Anesth
Reanim 1992;11:354-62.
18 Anderson DR, O'Brien BJ, Levine MN, Roberts R, Wells PS, Hirsh
J. Efficacy and cost of low-molecular-weight heparin compared with
standard heparin for the prevention of deep vein thrombosis after total
hip arthroplasty. Ann Intern Med 1993;119:1105-12.
19 Jorgensen LN, Wille-Jorgensen P, Hauch O. Prophylaxis of
postoperative thromboembolism with low molecular weight heparins.
Br J Surg 1993;80:689-704.
20 Mohr DN, Silverstein MD, Murtaugh PA, Harrison JM. Prophylactic
agents for venous thrombosis in elective hip surgery: meta-analysis of
studies using venographic assessment. Arch Intern Med
1993;153:2221-8.
21 Imperiale TF, Speroff T. A meta-analysis of methods to prevent
venous thromboembolism following total hip replacement.
JAMA 1994;271:1780-5.
22 Green D, Hirsh J, Heit J, Prins M, Davidson B, Lensing AW. Low
molecular weight heparin: a critical analysis of clinical trials.
Pharmacol Rev 1994;46:89-109.
23 Koch A, Bouges S, Ziegler S, Dinkel H, Daures JP, Victor N. Low
molecular weight heparin and unfractionated heparin in thrombosis
prophylaxis after major surgical intervention: update of previous
meta-analysis. Br J Surg 1997;84:750-9.
24 Borris LC, Lassen MR, Jensen HP, Andersen BS, Poulsen KA.
Perioperative thrombosis prophylaxis with low molecular weight heparins
in elective hip surgery: clinical and economic considerations.
Int J Clin Pharmacol Ther 1994;32:262-8.
25 O'Brien BJ, Anderson DR, Goeree R. Cost-effectiveness of
enoxaparin versus warfarin prophylaxis against deep-vein thrombosis
after total hip replacement. Can Med Assoc J
1994;150:1083-90.
26 Kearon C, Hirsch J. Starting prophylaxis for venous
thromboembolism postoperatively. Arch Intern Med
1995;155:366-72.
27 Beyth RJ, Landefeld CS. Anticoagulants in older patients: a
safety perspective. Drugs Aging 1995;6(1):45-54.
28 Borris LC, Lassen MR. A comparative review of the adverse
effect profiles of heparins and heparinoids. Drug Saf
1995;12(1):26-31.
29 Murray DW, Britton AR, Bulstrode CJK. Thromboprophylaxis and
death after total hip replacement. J Bone Joint Surg
1996;76B:863-70.
30 Andaz S, Shields A, Scurr JH, Coleridge-Smith PD. Role of low
molecular
31 Wells PS, Lensing AW, Hirsh J. Graduated compression stockings
in the prevention of postoperative venous thromboembolism: a
meta-analysis. Arch Intern Med 1994;154(1):67-72.
32 Clagett GP, Anderson FA Jr, Levine MN, Salzman EW, Wheeler HB.
Prevention of venous thromboembolism. Chest
1992;102(suppl 4):391-407S.
33 Goldhaber SZ, Morpurgo M. Diagnosis, treatment, and prevention
of pulmonary embolism. Report of the WHO/International Society and
Federation of Cardiology Task Force. JAMA
1992;268:1727-33.
34 Thromboembolic Risk Factors (THRIFT) Consensus Group. Risk of
and prophylaxis for venous thromboembolism in hospital patients.
BMJ 1992;305:567-74.
35 Haas S. European consensus statement on the prevention of
venous thromboembolism. Blood Coag Fibrin
1993;4(1):S5-8.
36 Skrabanek P. Nonsensus consensus. Lancet
1990;335:1446-7.
37 Conan Doyle A. Sherlock Holmes: the complete short stories.
London: John Murray, 1928.
38 Egger M, Davey Smith G. Misleading meta-analysis.
BMJ 1995;310:752-4.
(Accepted 26 August 1997)
Holmes's critical appraisal
Holmes sighed as he stood before us.
Holmes's conclusions
Holmes nodded. "Let us first confine ourselves to the
essential facts. Surgeons are concerned about safety and effectiveness,
and this has led to variations in practice as they seek to adapt
conflicting evidence to the circumstances of particular patients. Some
surgeons even believe that prophylaxis is not worth while, and that
pulmonary embolism is rarer than often suggested."
Dissemination,
University of York,
York YO1 5DD
research fellow
Susan C Kennedy,
adult basic skills
tutor
Home | Current issue | Past issues |
Classified ads | Career Focus | Feedback
Collections |
About this site | About the BMJ | BMA | Medline