Foci for evaluating accuracy of randomised and observational studies—simplified example for vitamin intake and cardiovascular outcomes
Baseline Demographic characteristics of patients (eg, age) Primary vs secondary prevention of disease Inclusion or exclusion of co-morbid conditions
Exposure Supplements vs dietary intake vs measured serum levels Dose and duration of therapy or co-therapy (as applicable) Pertinent confounding factors (eg, level of physical activity) Outcome Overall mortality vs cause-specific mortality vs morbid events Single vs combined endpoint (eg, non-fatal myocardial infarction or death) Timing (re start of exposure) and duration of follow-up
scientific quality, as well as the humanistic relevance, of patient-oriented research cannot be overstated. Our second proposal emphasises accuracy, precision, and repeatability when assessing study results, especially for therapeutic interventions. Accuracy refers to whether study results are true for those who might receive the intervention, incorporating concepts of both internal and external validity. Observational studies may have an overall advantage in this regard by typically including a broad population of participants,3 as problems with particular research procedures are identified and corrected.14–16 Precision refers to the level of numerical or clinical detail with which a result is generated. Observational studies may have an advantage in numerical detail, if only for being less expensive on a patient-for-patient basis, allowing for larger sample sizes; whereas randomised controlled trials can usually record clinical detail more readily. Repeatability refers to generating consistent results if similar investigations are conducted. For this criterion, randomised controlled trials have an advantage, with more explicit protocols for administering treatments and other interventions. With this conceptual approach, studies could be assessed systematically for accuracy, based on their baselineexposure-outcome framework (panel), but without a-priori pronouncements elevating a single randomised trial to gold-standard status.17,18 An intellectual shift in this direction appears to have begun already, as exemplified by recent assessments19,20 of hormone replacement therapy that avoid a prejudice against observational studies. Evaluations of precision and repeatability could also be done. Flaws in current approaches to patient-oriented research—with observational or randomised designs—are the basis for the cartoon included in the article by Vandenbroucke, and represent an unacceptable situation. Increased attention is needed to improve the basic science of patient-oriented research (clinical epidemiology), and to evaluate causal relations more rigorously. We have no conflict of interest to declare.
*John Concato, Ralph I Horwitz Clinical Epidemiology Research Center, West Haven Veterans Affairs Medical Center, West Haven, Connecticut; and Department of Medicine, Yale University School of Medicine, New Haven, CT 06510, USA (JC); and Office of the Dean, Case Western Reserve University School of Medicine, Cleveland, Ohio, USA (RIH) (e-mail: [email protected]
Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariable models. Ann Intern Med 1993; 118: 201–10.
THE LANCET • Vol 363 • May 22, 2004 • www.thelancet.com
16 17 18 19 20
Gunnell D. Can adult anthropometry be used as a ‘biomarker’ for prenatal and childhood exposures? Int J Epidemiol 2002; 31: 390–94. Concato J, Shah N, Horwitz RI. Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med 2000; 342: 1887–92. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med 2000; 342: 1878–86. McKee M, Britton A, Black N, McPherson K, Sanderson C, Bain C. Methods in health services research. Interpreting the evidence: choosing between randomised and non-randomised studies. BMJ 1999; 319: 312–15. Sacks H, Chalmers TC, Smith HJ. Ramdomized versus historical controls for clinical trials. Am J Med 1982; 72: 233–40. Kunz R, Oxman AD. The unpredictability paradox: review of empirical comparisons of randomised and non-randomised clinical trials. BMJ 1998; 317: 1185–90. Horwitz RI. Complexity and contradiction in clinical trial research. Am J Med 1987; 82: 498–510. Rabeneck L, Viscoli CM, Horwitz RI. Problems in the conduct and analysis of randomized clinical trials: are we getting the right answers to the wrong questions? Arch Intern Med 1992; 152: 507–12. Goodman SN. The mammography dilemma: a crisis for evidence-based medicine? Ann Intern Med 2002; 137: 363–64. Gotzsche PC, Olsen O. Is screening for breast cancer with mammography justifiable? Lancet 2000; 355: 129–34. Guyatt GH, Sackett DL, Sinclair JC, for the Evidence-Based Medicine Working Group. Users’ guides to the medical literature IX: a method for grading health care recommendations. JAMA 1995; 274: 1800–04. Erratum in JAMA 1996; 275: 1232. Feinstein AR, Horwitz RI. Problems in the “evidence” of “evidence-based medicine”. Am J Med 1997; 103: 529–35. Feinstein AR, Horwitz RI, Spitzer WO, Battista, RN. Coffee and pancreatic cancer: the problems of etiologic science and epidemiologic case-control research. JAMA 1981; 246: 957–61. Concato J, Horwitz RI, Feinstein AR, Elmore JG, Schiff SF. Problems of comorbidity in mortality after prostatectomy. JAMA 1992; 267: 1077–82. Concato J. Challenges in prognostic analysis. Cancer 2001; 91: 1607–14. U.S. Preventive Services Task Force. Guide to clinical preventive services, 2nd edn. Baltimore: Williams & Wilkins, 1996: 861–62. Pocock SJ, Elbourne DR. Randomized trials or observational tribulations? N Engl J Med 2000; 342: 1907–09. Grodstein F, Clarkson TB, Manson JE. Understanding the divergent data on postmenopausal therapy. N Engl J Med 2003; 348: 645–50. Garbe E, Suissa S. Hormone replacement therapy and acute coronary outcomes: methodological issues between randomized and observational studies. Hum Reprod 2004; 19: 8–13.
In-vitro model offers insight into the pathophysiology of severe malaria The main target against which control measures for malaria are directed is Plasmodium falciparum, the cause of some 3 million deaths and 500 million clinical episodes of malaria worldwide every year. The most serious complication of infection is cerebral malaria, in which the patient lapses into a coma from which they are often unrousable unless prompt treatment is given. Although the mechanisms are debated, cerebral malaria is associated with sequestration of infected erythrocytes, the stage of the malaria life-cycle that is responsible for the pathogenesis of the disease, in the microvasculature of the brain. Excessive accumulation of both infected and uninfected erythrocytes causes a mechanical blockage that leads to cessation of the local blood supply. This effect is widely assumed to decrease the availability of oxygen to neurons, thus contributing to the neurological abnormalities that ensue.1 Parasite sequestration itself is due to binding of infected erythrocytes to the vascular endothelium (cytoadherence) and to surrounding uninfected erythrocytes (rosetting), through several specific receptors.2 Although a fatal outcome of falciparum malaria is most commonly linked to sequestration to the brain, blockage of deep-tissue capillaries also causes problems in other organs, especially the spleen, liver, kidney, and placenta. 1661
For personal use. Only reproduce with permission from The Lancet Publishing Group.
Representation of blood flow through microchannel Although 8 m in width, normal, uninfected erythrocytes are flexible enough to pass through central constriction of 2 m diameter. By contrast, when infected with P falciparum, erythrocytes become spheroid and decreasingly deformable, such that, as shown, they block 6 m constriction.
The severity of disease is thought to be directly proportional to the loss of native deformability of infected erythrocytes and thereby to the degree of vessel occlusion.3 Normal erythrocytes are extremely deformable, due to a combination of low internal viscosity, elasticity of their membrane and underlying cytoskeleton, and the high ratio of surface area to volume of their biconcave disc shape.4 However, during parasite development, malariaparasitised erythrocytes become increasingly spheroid, develop knob-like protrusions, and acquire rigidity. Because we lack suitable models, our knowledge of how these processes occur at the single-cell level and how they combine to produce pathological sequelae in vivo is fragmented. Although mammalian models have more than proved their worth in immunological and chemotherapy studies, the techniques required to make them a practical proposition for accurately investigating blood rheology in vivo are currently lacking. Pathological examinations of P falciparum in human brain can only take place during postmortem biopsies,5 which are infrequent and from which is gleaned an understanding of just one snapshot of infection at the time of death but not of the preceding stages. Malaria researchers have, therefore, over the past two decades addressed many of the issues surrounding the pathophysiology of P falciparum by investigating infected erythrocytes cultivated in vitro. This research has greatly increased knowledge of the molecular mechanisms of cytoadherence and rosetting, but has failed to answer how parasite sequestration leading to capillary blockage occurs in the body as a whole. What has been missing is a system that closely mimics the in-vivo environment. Patrick Shelby and colleagues6 have reported the development of an in-vitro model that effectively simulates capillary occlusion by erythrocytes infected by P falciparum, right down to blockage by a single cell maintained under conditions of physiological blood flow. Although previous in-vitro rheological techniques have informed understanding of erythrocyte deformability, they have been limited to the measurement of a single physical variable of a single cell or applied to bulk populations of cells.7 Shelby and colleagues solved this problem by the use of a microchannel device similar to that first used to measure the area, volume, and deformability of individual normal erythrocytes.8 Made from a silicone-based elastomer, microchannels were constructed to mimic average capillaries, having a similar elastic modulus and a typical bore size of between 2 and 8 m. The flow velocity of aspirated blood was regulated to model that in human capillaries in vivo, typically 100–500 m/s.9 Maintained in a controlled incubation environment, this system provides an accurate means to study the behaviour of malariaparasitised erythrocytes in conditions that as near as 1662
possible replicate intracapillary flow. The microchannel is illustrated in the figure. Shelby and colleagues6 show that as an infected erythrocyte matures, it becomes increasingly less able to traverse small-bore capillaries. Hence, newly infected erythrocytes, containing so-called ring-stage parasites, are as equally elastic as 8-m-wide uninfected erythrocytes, both passing freely through 2 m channels. However, the most mature stage of intraerythrocytic parasite, the schizont forms, fail to negotiate even channels of 6 m width. When capillaries get blocked as a result, uninfected erythrocytes and those recently infected, because of their greater deformability, can still squeeze through, which has the effect of concentrating mature infected erythrocytes at the site of a capillary obstruction. This observation may explain not only the high number of parasites in growing capillary blockages but also why severely reduced erythrocyte deformability is a strong predictor of mortality. The benefits of early transfusion of blood in individuals presenting with severe malaria would thus appear to be due to alleviation of acquired rigidity as well as anaemia. Shelby’s model6 is important because it offers the opportunity to examine properties of individual infected erythrocytes in real time in a carefully controlled capillarylike microenvironment. Although this first observation of erythrocytes infected by P falciparum under conditions of controlled flow-rate, temperature, and pressure is notable in itself, it also proves a point of principle by showing the robustness of the model. The microchannel device can be modified in several ways to change the elastic modulus and silicone surface-chemistry, and, because of its high gaspermeability, to enable the growth of endothelial cells along its inner surface. Detailed characterisation under physiological conditions of the way in which infected erythrocytes adhere to multiple receptors on the endothelium of capillaries should now be possible and is the next step forward. Such studies into the sequestration of P falciparum will lead to a better understanding of the pathogenesis of cerebral malaria and will aid the development of drugs and vaccines for treatment and prevention. The model therefore provides a powerful tool that could become as central to malaria research as the original technique for successful in-vitro cultivation of blood-stage P falciparum,10 with which it is intrinsically harnessed. I have no conflict of interest to declare.
Andrew Taylor-Robinson School of Biology, University of Leeds, Leeds LS2 9JT, UK (e-mail: [email protected]
) 1 2
Clark IA, Cowden WB. The pathophysiology of falciparum malaria. Pharmacol Ther 2003; 99: 221–60. Sherman IW, Eda S, Winograd E. Cytoadherence and sequestration in Plasmodium falciparum: defining the ties that bind. Microbes Infect 2003; 5: 897–909. Dondorp AM, Kager PA, Vreeken J, White NJ. Abnormal blood flow and red blood cell deformability in severe malaria. Parasitol Today 2000; 16: 228–32. Cooke BM, Mohandas N, Coppel RL. The malaria-infected red blood cell: structural and functional changes. Adv Parasitol 2001; 50: 1–86. Grau GE, Mackenzie CD, Carr RA, et al. Platelet accumulation in brain microvessels in fatal pediatric cerebral malaria. J Infect Dis 2003; 187: 461–66. Shelby JP, White J, Ganesan K, Rathod PK, Chiu DT. A microfluidic model for single-cell capillary obstruction by Plasmodium falciparuminfected erythrocytes. Proc Natl Acad Sci USA 2003; 100: 14618–22. Cooke BM, Coppel RL, Nash GB. Analysis of the adhesive properties of Plasmodium falciparum-infected red blood cells under conditions of flow. Methods Mol Med 2002; 72: 561–69. Gifford SC, Frank MG, Derganc J, et al. Parallel microchannel-based measurements of individual erythrocyte areas and volumes. Biophys J 2003; 84: 623–33. Shelby JT, Chiu DT. Mapping fast flows over micrometer-length scales
THE LANCET • Vol 363 • May 22, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
using flow-tagging velocimetry and single-molecule detection. Anal Chem 2003; 75: 1387–92. 10 Trager W, Jensen JB. Continuous culture of Plasmodium falciparum: its impact on malaria research. Int J Parasitol 1997; 27: 989–1006.
Tongue transplantation On July 19, 2003, a Viennese team of surgeons did the world’s first transplantation of a tongue, grafting a tongue from an ABO-matched heart-beating donor into a 42-year-old man with tongue cancer. The recipient was discharged from hospital a month later with a tracheostomy for airway support and a gastrostomy for nutrition. A hospital spokesperson told the press “Although the tongue still can’t be moved, the patient can already swallow some of his own saliva,” and that he “was already making himself understood”. The graft showed no signs of rejection (8 months’ follow-up to the end of March), and has some useful sensation, enabling the patient to swallow all saliva and some fluids, although some muscle wasting is occurring. When asked about the long-term prospects, the surgeon, Rolf Ewer said: “The liver and kidneys are complicated organs, but the tongue is just a muscle so it should work out.” However, it was thought unlikely that the sense of taste would return.1,2 As Day3 has put it, “the tongue is not a vital organ in sustaining life, [but] it may be a vital organ in sustaining the will to live in many people.” Head and neck surgeons agree that advanced cancer of the tongue, particularly at its base, is one of the most difficult problems in head and neck surgery. The only previously published cases of tongue replantation in reality only involved the reattachment of the anterior portion of a human tongue after physical trauma.4 It might be expected that immunosuppression would lead to a higher rate of recurrent head and neck cancer than would be seen without such treatment. However, surprisingly, the rates squamous cell cancer in the head and neck in big series of immunosuppressed patients are no different from those of population controls.5,6 Nonetheless, reconstruction without transplantation after total glossectomy produces unpredictable speech and swallowing results,7 and recurrence rates for local cancer are high. Therefore many are watching the long-term results of the Viennese operation with great interest. This case represents only the latest in a recent wave of transplantations for quality of life rather than quantity. On the assumption that the morbidity of modern immunosuppressive regimens is less than that of life without a functional larynx, hand, tongue, or face, transplantation of the first three organs has now been done with some success,8–10 and face transplants have recently encouraged debate.11 In fields where conventional medicine or technological advances have made little impact, these operations suggest the possibility of restoring “humanity” to those with no voice, swallowing, or ability to touch and manipulate the environment, or to those severely facially disfigured. Tongue, laryngeal, hand, and face transplantations share three other features: they all excite a great deal of lay (and mass media) interest and thus publicity; they apply (initially at least) to only tiny numbers of potential recipients; and, with the possible exception of laryngeal transplantation, they have all been the subject of very little preclinical (especially immunological) research before attempted translation into human beings. The amount of publicity generated leads, rightly or wrongly, to scepticism about the motives of those surgeons who pioneer these operations. Such surgeons need to make the case for THE LANCET • Vol 363 • May 22, 2004 • www.thelancet.com
transplantation crystal clear, to convince others of the ethical and clinical rationale. The pool of potential recipients needs careful assessment, even if a successful pilot series may lead to a relaxing of inclusion criteria with larger numbers later on. Clearly the introduction of any of these quality-of-life transplantations must be preceded by an adequate period of preclinical research. Such research should include consideration of: what, how much, and indeed whether immunosuppression should be used; the effects of ischaemia-reperfusion injury; and whether we can provide truly functional reinnervation (a fundamental question in all these examples). This research might be seen as an unnecessary and expensive delay to the introduction of techniques that may rehumanise some unfortunate patients. However, if we cannot prove that these organs can function normally in a valid preclinical model, should we really progress to human trials in which the only certain outcome is publicity? I thank Rolf Ewers for updates about the patient. I have no conflict of interest to declare.
Martin Birchall Head and Neck Cancer Centre, University Hospital Aintree, Liverpool L7AL, UK (e-mail [email protected]
BBC News. Tongue transplant patient doing well. July 22, 2003: http://news.bbc.co.uk/1/hi/health/3086509.stm (accessed March 17, 2004). 2 Anonymous. Recipient of first tongue transplant recovering. CDS Rev 2003; 96: 44. 3 Day TA, Tarr C, Zealear D, Burkey BB, Sullivan CA. Tongue replantation in an animal model. Microsurgery 2000; 20: 105–08. 4 Buntic RF, Buncke HJ. Successful replantation of an amputated tongue. Plast Reconstr Surg 1998; 101: 1604–07. 5 Mihalov ML, Gattuso P, Abraham K, Holmes EW, Reddy V. Incidence of post-transplant malignancy among 674 solid-organtransplant recipients at a single center. Clin Transplant 1996; 10: 248–55. 6 Marcen R, Pascual J, Tato AM, et al. Influence of immunosuppression on the prevalence of cancer after kidney transplantation. Transplant Proc 2003; 35: 1714–16. 7 Zuydam AC, Rogers SN, Brown JS, Vaughan ED, Magennis P. Swallowing rehabilitation after oro-pharyngeal resection for squamous cell carcinoma. Br J Oral Maxillofac Surg 2000; 38: 513–18. 8 Strome M, Stein J, Esclamado R, et al. Laryngeal transplantation and 40-month follow-up. N Engl J Med 2001; 344: 1676–79. 9 Dubernard JM, Petruzzo P, Lanzetta M, et al. Functional results of the first human double-hand transplantation. Ann Surg 2003; 238: 128–36. 10 Klotzko AJ. Willing hands . . . but new nonvital transplants could be raising more emotional problems than they solve. New Sci 1999; 162: 51. 11 Hettiaratchy S, Butler PE. Face transplantation—fantasy or the future? Lancet 2002; 360: 5–6.
Rape of individuals with disability: AIDS and the folk belief of virgin cleansing Virgin cleansing—the belief that people who have a sexually transmitted disease can rid themselves of the condition by transferring the infective organism by having sexual intercourse with a virgin—has been discussed in The Lancet in relation to HIV/AIDS.1–3 The practice was first reported in the 16th century in relation to syphilis and gonorrhoea in Europe.4 Although the prevalence of virgin cleansing is unclear,1–3 accounts of the belief are reported from sub-Saharan Africa, Asia, Europe, and the Americas.5 We have identified a variation of this practice in our Global Survey on HIV/AIDS and Disability6 that warrants attention—“virgin rape” of individuals with disability, by 1663
For personal use. Only reproduce with permission from The Lancet Publishing Group.