1441 The inventory can be used for individuals, groups of students, or clinicians at all levels. It can also be used to assess the effectiveness of education and training in developing clinical reasoning. The diagnostic thinking inventory is content-free and can be used in any specialty. With such a validated instrument already available, the state of the art is perhaps more advanced than van der Vleuten and Newble suggest. *Janet Grant, Philip Marsden *Joint Centre for Education in Medicine, London SE10 9AG; and Greenwich District Hospital, London WC1N 3EJ, UK 1 Bordage G, Grant J, Marsden P. Quantitative assessment of diagnostic ability? Med Educ 1990; 245: 413-25. SIR—Godfrey comments (March 25, p 765) that the dry drudgery of undergraduate examination details is tending to produce paralysing mental sterility in medical students. The fashion of accepting for medical school admission only students with straight A grades in college excludes many inquiring and innovative minds. I have been in the forefront of new medical concepts and techniques, and this loss is evident to me. I find that young doctors who have achieved straight A grades in college by virtue of their ability to repeat verbatim what they are taught are resistant to new ideas and approaches. George E Shambaugh Shambaugh Hearing and Allergy, PC, Hinsdale, IL 60521, USA SiR-Your thoughtful series on medical education consists of an introductory commentary and seven articles (March 25, to May 6). It would be impertinent to attempt proper critique in a short letter, so here is a brief shot and one comment. The consultation is the centre piece of medical practice, in which the patient rightly expects the doctor to pass in, say, a 20 minute unprepared encounter. Clinical final MB examinations based on this model have a large claim to a real-life test. The comment is from Rose Macaulay.’ I have adulterated the extract with brackets. She is talking about 17th century religion (medical education), and says "... many parsons [deans] conformed to the new order, while many sequestered divines still administered the sacraments in the Anglican style ... not permitting themselves to be thrown out of their stride ... even by the rather indecorous behaviour of their congregations. For there were often strange seizures in church [medical school], sudden undressings and leapings, spasmodic and piercing cries of devotion and repentance, such as are apt to obtain among devout persons moved beyond endurance by the spirit. Possibly the new drinks went to people’s heads, for, though strong liquors were less in vogue, tea, coffee and chocolate [OSCE] began to be drunk, and these, accompanied by ceaseless religious and political discussion [medical education conferences], were unsteadying. So was constant newspaper reading [The Lancet], which now became a permanent British vice; in coffee houses [curriculum committees] they passed the news sheets round, read them out, and passionately debated them. So much sectarian zeal was heating". T Sherwood University of Cambridge School of Clinical Medicine, Addenbrooke’s Hospital, Cambridge CB2 2SP, UK 1 Macaulay R. Life among the English. London: William Collins, 1942: 30. Serum HIV-1 RNA and routine patient monitoring SiR-The most common approach to management of HIV- infected patients without AIDS is to monitor both the CD4 lymphocyte count and the development of early HIV- related symptoms every 3-6 months to assess whether the patient is at high risk of developing AIDS before their next clinic visit. If this is found to be the case-most commonly because the CD4 count has fallen below 200/L—the patient may be offered therapy. Loveday and Hill (March 25, p 790) report that the HIV RNA concentration provides extra information, in addition to that provided by the CD4 count, about whether a patient will develop AIDS. If their results are replicated the obvious implication would be that plasma HIV RNA concentration should be used in routine patient monitoring; this, however, would not necessarily be so. Patients with high plasma HIV RNA may have developed AIDS more rapidly over the ensuing 2 years than patients with similar CD4 counts but lower HIV RNA concentrations because of a more rapid decline in CD4 count during this period. It is noteworthy that there are very large differences in risk of AIDS according to CD4 count, even among patients with counts below 200/[tL,’ so the rate of CD4 count decline during follow-up is important even for patients with initially low counts. If so, as long as patients are being seen every few months, regular monitoring of CD4 counts may suffice. Since patients in Loveday and Hill’s study probably had regular CD4 count monitoring during the 2 years of the study, it would be interesting to consider the size of the association between plasma HIV RNA and risk of AIDS after adjustment for the CD4 count as a time-updated (ie, taking account of new counts during follow-up as they are made) variable in the Cox proportional hazards model. If an association of appreciable magnitude remains after such adjustment then this would suggest that plasma HIV RNA should indeed be used in patient monitoring. If the association is substantially attenuated after the adjustment then-since we assume that HIV RNA values were not updated over time in Loveday and Hill’s study-the issue can only be resolved in new studies in which this is done. *Andrew N Phillips, Caroline A Sabin, Amanda Mocroft University Department of Public Health, Royal Free Hospital School of Medicine, London NW3 2PF, UK 1 Phillips AN, Pezzotti P, Cozzi Lepri A, Rezza G, and the Italian Seroconversion Study. CD4 lymphocyte count as a determinant of the time from HIV seroconversion to AIDS and death from AIDS: evidence from the Italian Seroconversion Study. AIDS 1994; 8: 1299-305. SiR-Loveday and Hill’s data for viral load and the relation to the emergence of zidovudine-resistant genotypes indicated that early viral escape from zidovudine is independent of the emergence of genetic viral resistance. As they noted, a portion of this early viral escape in the presence of zidovudine may be secondary to "altered drug activity at one compartment or site". We have described an analysis of early viral escape from zidovudine in a cell-culture model that is directly relevant to these conclusions.’ Like those presented by Loveday and Hill, we recorded early viral escape by genetically sensitive virus. In addition, we established that a component of this viral escape occurs in a subset of cells in which zidovudine is an ineffective antiviral agent. We have initiated a biochemical analysis of some of these cells and have shown alterations in zidovudine triphosphate accumulation and thymidine phosphorylation as potential factors contributing to the inefficacy of zidovudine in these cells.