A 1998 study joined baseline testing, controls and follow-up.
Reading Library · Anesthesia Epochs · 中文版
Watch the companion film
1998: ISPOCD1 — Postoperative Cognitive Dysfunction · Watch on YouTube
Historical anchor: 1998. Date details and source qualifications are explained below.

Leaving the operating room does not answer every question about recovery. In ISPOCD1, published on March 21, 1998, Moller and colleagues followed cognitive performance after major noncardiac surgery under general anaesthesia. Thirteen hospitals enrolled 1,218 patients aged 60 years or older.[1, pp. 857–859] This was a prospective multicentre study. Anaesthetic and surgical techniques followed local practice; anaesthetic exposure was not randomly assigned.
Measuring a decline required more than a low postoperative score. Investigators needed a baseline and a way to account for improvement through practice. Testing was scheduled before surgery, at discharge or one week after surgery, whichever came first, and at three months. Patients too ill for the early assessment could be tested later. The final assessment used four tests. Changes in 176 UK volunteers who did not undergo surgery supplied the correction for learning effects; another 145 national controls helped assess whether the method could be applied across the participating populations.[1, pp. 858–859]

The tests covered visual verbal learning, concept shifting, Stroop colour-word interference and letter-digit coding. The investigators converted changes into Z scores and applied their study definition of postoperative cognitive dysfunction, or POCD. This was not a diagnosis of dementia based on a single MMSE screen. Nor was it interchangeable with delirium: the orientation component of the MMSE was used separately each day to screen for confusion and delirium.[1, p. 859]
At the corresponding postoperative assessments, 266 patients, or 25.8%, met the POCD definition at about one week, and 94, or 9.9%, at three months. The corresponding UK control proportions were 3.4% and 2.8%.[1, pp. 857, 859] These figures describe the study's measured outcome among those assessed. They do not mean that every enrolled patient completed both follow-ups. Of the 1,218 patients enrolled, 271 did not complete the three-month assessment. The three-month risk-factor regression used a further distinct denominator of 910.[1, pp. 858–859]

There is also a correction to read alongside the paper. On June 6, 1998, The Lancet published a notice on page 1742 changing thirteen cells in Table 3. For example, the three-month entry for patients aged 70 or older changed from “4 (14%)” to “52 (14%).”[2] The overall findings described here were checked against this notice. Its stated scope is Table 3; it does not resolve every issue elsewhere in the original tables.
Interpretation required another distinction. Age was associated with POCD at three months, but this observational comparison could not isolate the effect of a particular anaesthetic drug. The discussion left mechanism, permanence and applicability to younger patients or minor operations open to further study. The authors also stated that they could not infer safe degrees of hypoxaemia or hypotension.[1, pp. 859–860] The 9.9% figure is therefore neither a universal rate for surgical patients nor a rate of permanent dementia caused by anaesthesia.
The later 2018: The PND Consensus Separates Postoperative Time from Cause explicitly cited ISPOCD1. They described the difficulties created by differences in tests, definitions and assessment times, while bringing subjective cognitive concerns and everyday function into a clinical terminology framework.[3, pp. 1006, 1011] A 2019 brain-health review also cited the ISPOCD figure of 25.8%, but distinguished research-defined POCD from newer diagnostic terms requiring additional clinical criteria.[4, PDF pp. 2, 9] These papers document how the study remained part of the discussion. Their citations are not independent replications of its patient results.

ISPOCD1 brought baseline assessment, a learning-effect control and follow-up beyond discharge into one research design. That widened the window through which recovery could be observed. Cognitive change measured within that window still needed to be distinguished from an explanation of its cause and from a clinical diagnosis.[1, pp. 858–860; 3, p. 1006]
All four illustrations are modern editorial reconstructions or concepts, not photographs of the study, original test sheets, patient records or procedural demonstrations.
Sources and reading limits
- Moller JT et al. Long-term postoperative cognitive dysfunction in the elderly: ISPOCD1 study. Lancet. 1998;351:857–861. The Maastricht publisher PDF was read in full: its repository cover and all five original article pages, in text and original page images.
- Department of Error: the ISPOCD1 notice. Lancet. 1998-06-06;351(9117):1742. The complete middle notice was read on the original scan. No individual DOI for the notice could be confirmed.
- Evered L et al. Recommendations for the nomenclature of cognitive change associated with anaesthesia and surgery—2018. Br J Anaesth. 2018;121:1005–1012. Only pp. 1006 and 1011 were read in text and original images.
- Mahanna-Gabrielli E et al. State of the clinical science of perioperative brain health. 2019. Only pages 1, 2 and 9 of the ASA-hosted pre-corrigendum article-in-press copy were read: identity, relevant paragraphs and reference 11. The old byline is not presented as the final corrected author list.
The original Table 1 sex counts and the country total versus the end-of-paper list remain discrepant and are not used here. Later consensus and review articles depend on the original research; they do not supply three independent experiments.
Sources and reading limits are stated above. Illustrations are labelled as original source images or AI-made. Historical accounts are educational; they are not current treatment instructions.
Continue exploring
Related reading
Related reading connects shared questions and contexts; it does not by itself establish historical influence.