n=1 or n=megabytes?

I had thought there were two types of SHOs (residents): those who analysed clinical problems, considered solutions, then requested approval; and those who placed their intellect carefully in neutral and asked what to do. Post Sackett,¹ a third type is emerging—the cyber-SHO, armed with fingers as sensitive to a palmtop keyboard as to a 25 g Venflon.

We invited one such to let us into his professional secrets (page 373). The trouble is I'm not sure I agree with the evidence-based solution to his clinical vignette.

 1 Sackett DL, Strauss SE, Richardson WS, et al. Guidelines. In: Evidence-based medicine: how to practise and teach EBM. Edinburgh: Churchill Livingstone, 1996.

n = blind prejudice

In contrast, we also look at habit-based medicine. Leeds (UK) is fortunate in having a department of clinical effectiveness. (Does anyone have a department of clinical ineffectiveness?) Farshi and colleagues (page 393) surveyed local hospital paediatricians and surgeons on their attitudes to circumcision. Unsurprisingly, surgeons tended towards cutting and physicians to preserving. Individual opinions varied greatly; alarmingly, 15% of surgeons approved of operating on children under 5 with painless non-retractile foreskins. More alarmingly, 60% of paediatricians did not know what to do about balanitis xerotica obliterans—perhaps the only absolute surgical indication. For those prepared to relearn old lessons, see Gairdner D. The fate of the foreskin. BMJ1949:2;1433–7. For those preferring them when they've come round a few times, try American Academy of Pediatrics Task Force on Circumcision. Circumcision policy statement.Pediatrics1999;103:686–93.

n = grand old boys + rigour

ADC is cautious, indeed suspicious, of working party reports as they may circumvent the normal peer review process. This month we have thrown caution to the winds in publishing details of guidelines for the treatment of epileptic status, as recommended by an ad hoc group (page 415). Our reason for heterodoxy is approval of the process: when a properly conducted systematic review found only two randomised controlled trials, the other 371 papers containing original data were graded according to level of evidence and only then submitted to consensus. Moreover, the group tells us they have designed two multicentre RCTs looking at one step in their guidelines as well as auditing the whole. We look forward to being invited to publish the results.

n = fewer than you think

Blind broad spectrum antibiotic prescribing bothers all of us, whether it arises from uncertainty, defensiveness, or ignorance. Conditions such as community acquired pneumonia (CAP) invite all three because of the difficulty of making an aetiological diagnosis. Studies from the USA placed a spotlight on Mycoplasma pneumoniae ²; a Finnish group has incriminatedStreptococcus pneumoniae in more than one third of inpatients.³ Published studies vary in diagnostic criteria, case mix, and modality of testing, so incompatible results are unsurprising.  This month, a hospital study from Newcastle (UK) investigated 136 children admitted with CAP (page 408). Despite the investigators' efforts, half remained unclassified; in the remainder, viruses (mainly RSV) were causative in 71%. The authors' main conclusion was that in children under 2 years old,S pneumoniae might be responsible for one in five cases. The study was not designed to produce treatment recommendations, and the authors advise on the need for more rigorous serological testing and for investigation of the cost effectiveness of pneumococcal vaccines.  Serendipitously, Toikka and colleagues from Turku (Finland) complicate the issue by describing nine cases ofM pneumoniae and S pneumoniae coinfection (page 413).

 2 Denny FW, Floyd WA. Acute lower respiratory tract infections in non-hospitalised children. J Pediatr1986;108:635–46.

 3 Juven T, Mertsola J, Waris M, et al. Etiology of community-acquired pneumonia in 254 hospitalised children. Pediatr Infect Dis J2000;19:293–8.

n = another small step for (small) mankind

More on the postural effects on infant physiology (page423). For the first time (we think), data have been collected on ventilatory and arousal responses of 3 month olds to a mild asphyxial challenge. Prone positioning reduced the ventilatory response but increased arousal in active sleep to 13.5% oxygen. This further paragraph in the SIDS story from New Zealand complements the authors' observations on head up tilt testing.⁴ The authors speculate that babies vulnerable to SIDS may fail to respond and arouse to the various stressors related to prone lying. By the way, given current anxieties, informed consent was obtained and no baby developed a low SaO₂ or a clinically significant heart rate change—evidence for the safety of the research protocol.

 4 Galland BC, Reeves G, Taylor BJ, et al. Sleep position, autonomic function and arousal. Arch Dis Child 1998;78:F189–94.

n = heads or tails

Kerr and colleagues produce interesting histopathological and DNA extraction evidence for an association between SIDS andH pylori infection (page 429). As this edition went to press, we received a paper concluding that there is no such relationship. We eagerly await the end of the peer review process.

n = blissful amnesia

Another possible first is to ask children what they recall of a stay in an intensive care unit. Playfor and colleagues from Nottingham (UK) interviewed 38 of 45 children admitted to their paediatric intensive care unit (page 445). Most remembered being there and half knew why; thankfully specific memories were mostly neutral or positive. Three complained about noise, one about not knowing what time of day it was, two about nasogastric tubes, and one about his endotracheal tube. One third of the children recalled pain but there were almost no mentions of ventilation or paralysis. The authors give thanks to midazolam.