The Informant Questionnaire on Cognitive Decline in the Elderly – Short Form (IQCODE-16; Jorm, 1994) is a 16-item informant questionnaire about cognitive decline and possible dementia in older adults (65+).
The Informant Questionnaire on Cognitive Decline in the Elderly – Short Form (IQCODE-16; Jorm, 1994) is a 16-item questionnaire about cognitive decline and possible dementia in an older adult, completed by a relative or friend who has known the person for 10 years or more (Australian National University, n.d.). The respondent compares the person now with 10 years ago on everyday tasks that rely on memory and thinking, such as recalling recent events, handling money, and learning to use a new gadget (Jorm, n.d.).
The IQCODE-16 has been used with adults aged 65 and over, in the community and in hospital (Harwood et al., 1997; Jorm, 1994). The version implemented on NovoPsych is the 16-item English short form, which its publisher calls the Short IQCODE and the author recommends as the preferred English version (Jorm, 2004).
Cognitive decline is a reduction in a person’s memory and thinking from their own earlier level, and the IQCODE measures this change (Jorm, 1994). Because it asks about change, ratings are less tied to a person’s education and earlier ability than cognitive test scores (Jorm, 1994).
The IQCODE-16 is a screen for dementia in older adults, used to decide who needs a fuller assessment (Quinn et al., 2021). It is not a diagnostic test. It can be given alongside a cognitive test. When both the IQCODE and the Mini-Mental State Examination (MMSE) are given, the author recommends investigating further anyone whose result suggests a problem on either test (Jorm, 2004).
Because an informant completes it, the questionnaire can be used when the person is too unwell or otherwise unable to complete a cognitive test (Harwood et al., 1997). It also adds detail to the history and formulation. Each item asks about a specific everyday ability, such as handling money or recalling recent conversations, so the answers show where the informant has noticed change (Jorm, n.d.).
The IQCODE-16 can be sent straight to the informant. The User Guide shows how to send assessments to informants.
The IQCODE-16 presents a score between 1 and 5, which is the average of the 16 ratings (Australian National University, n.d.). The IQCODE-16 has no norms or percentiles, so the average is read against the publisher’s descriptions and the screening threshold. Higher averages mean the informant reports more decline.
The answers are listed in three groups: rated as worse (most reported decline first), not much change, and rated as improved (most improvement first), with equal ratings in questionnaire order. The interpretive text names the items rated “Much worse” if there are five or fewer. The report opens with a note when the informant has known the client for less than the assumed 10 years (Australian National University, n.d.).
The Description column shows where the average sits against the screening threshold of 3.38. It reads “Above screening threshold”, shaded yellow, at 3.38 or more, and “Below screening threshold” below 3.38.
The ratings reflect the informant’s view, which their own mood and how they see their relationship with the client can influence. Depression in the client was linked to higher averages, and in one study so were the informant’s anxiety, low mood, and view of the relationship (Jorm, 1994; Jorm et al., 1996).
In memory clinics, pooled across both forms and a range of cut-offs, about half of the people without dementia scored above the cut-offs used. The review authors suggest that depression, common in these clinics, may contribute (Burton, Fearon, et al., 2021).
On a first administration, one chart shows the average on the 1 to 5 scale, with each number on the axis labelled in words, from “Much improved” at 1, through “No change” at 3, to “Much worse” at 5. A yellow band shades the area at or above the screening threshold of 3.38. The two areas are labelled “Above screening threshold” and “Below screening threshold” on the right of the chart.
Each rating compares the client with 10 years before the date it was completed. Repeat ratings appear on a line chart with the same yellow band and labels. Repeat ratings may come from different informants.
Jorm (1994) chose the 16 items of the IQCODE-16 from the 26-item form (Jorm & Jacomb, 1989), using four Australian datasets, and dropped items with low reliability or validity, or that were affected by education. The IQCODE-16 was then tested in a new survey of people aged 70 and over, living in the community or in residential care, of whom 684 had enough data for a computer-scored diagnosis (Jorm, 1994). The publisher offers three English forms: the original 26-item form, the IQCODE-16 short form (as implemented on NovoPsych), and a retrospective form (Australian National University, n.d.). Across five analyses, mostly of the 26-item form, a general factor explained 42% to 61% of the variance (Jorm, 2004).
In the studies reviewed, total-score reliability is reported only for the 26-item form: internal consistency was strong (alpha = .95; Jorm & Jacomb, 1989, 613 general population informants). The 26-item total had a retest correlation of .96 over an average of 3.2 days, in informants of geriatric and psychogeriatric patients (n = 57; Jorm et al., 1991). The IQCODE-16 total correlated .98 with the 26-item total (Jorm, 1994), a figure raised by the 16 items being part of the 26.
In the cross-validation sample, the IQCODE-16 correlated −.58 with the MMSE and .34 with age, and distinguished people with dementia or delirium from those without about as well as the MMSE (AUC .85 against .90, a non-significant difference; Jorm, 1994). The Cochrane reviews pooled studies of both forms and several languages, including some of the studies above. Across 11 community datasets of both forms, sensitivity was .80 and specificity .85, using each study’s cut-off nearest to 3.3 (from 3.3 to 3.6; Quinn et al., 2021). In secondary care, with both forms and cut-offs from 3.3 to 4.1 pooled, sensitivity was .91 and specificity .66, and specificity was lower in memory clinics (.54) than in general hospitals (.81; Burton, Fearon, et al., 2021).
The IQCODE-16 cut-off is between 3.31 and 3.38, the point at which sensitivity and specificity were balanced in the cross-validation sample (sensitivity .79, specificity .82; Jorm, 1994). The cut-off was set against a diagnosis of dementia or delirium in 52 cases among 684 people (Jorm, 1994). Two later studies applied this cut-off to the short form in new samples. In 144 ex-servicemen rated by their wives, it gave sensitivity .75 and specificity .68 against 24 cases diagnosed by a neurologist, and .91 and .65 against 11 cases diagnosed by a psychiatrist (Jorm et al., 1996). In 177 older medical inpatients, an average above 3.31 (in practice 3.38 or more) gave sensitivity of 100% and specificity of 78% against dementia (Harwood et al., 1997). The paper does not say how many of the 177 inpatients had dementia; NovoPsych estimates 24 to 26 from its tables. NovoPsych uses 3.38 as the screening threshold on the strength of the original study and these two smaller ones.
The questionnaire is designed for a relative or friend who has known the person for at least 10 years (Australian National University, n.d.). In the original research on the full questionnaire, the informant’s age and sex, and how long it was since they had lived with the person, made little difference to the ratings (Jorm & Jacomb, 1989). There was no evidence that older informants gave less valid ratings than younger ones (Jorm et al., 1991). What can make a difference is the informant’s own state of mind: in one study, wives who were more anxious or low in mood rated more decline in their husbands (Jorm et al., 1996).
In the original research, the full questionnaire was found not to be biased against people with less education (Jorm & Jacomb, 1989). When the short form was developed, the items most affected by education were dropped (Jorm, 1994).
Yes. The screening threshold was set against dementia or delirium, so a person with delirium may score above it. In one study of the full questionnaire, the ratings of four people with delirium changed more over a few days than other people’s did, which the authors put down to real improvement as the delirium settled (Jorm et al., 1991).
It can, but each rating looks back 10 years from the day it is completed, so two ratings a few months apart cover almost the same period. NovoPsych therefore shows each repeat rating as its own point on a line chart, rather than calculating how much the score has changed.
The author’s review suggests a higher cut-off of 3.44 for clinical settings (Jorm, 2004). The only short-form study at that cut-off in the review is one study of older medical inpatients, where an average above 3.44 (in practice 3.5 or more) identified all the patients found to have dementia, and flagged fewer of the patients without it than the screening threshold did (Harwood et al., 1997). Cut-offs published for the full 26-item questionnaire, such as 3.60, were set on that form, so NovoPsych does not apply them to the short form. NovoPsych’s screening threshold is the cut-off set when the short form was developed, which is also the one the publisher gives.
In the Cochrane review, three studies looked at this, two of them using translated short forms. In an Italian study of people who had had a stroke, an average above 3.3 identified most of those diagnosed with dementia a year later and few of those who were not. In a Norwegian study of people with a hip fracture, a lower threshold identified most of the 10 people diagnosed with dementia six months later, but also almost half of the people who were not. The Cochrane review authors concluded that the IQCODE was sensitive but not specific for later dementia and made no recommendation for this use (Burton, Stott, et al., 2021), so NovoPsych shows no threshold for predicting dementia.
Jorm, A. F. (1994). A short form of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): Development and cross-validation. Psychological Medicine, 24(1), 145–153. https://doi.org/10.1017/S003329170002691X
Australian National University. (n.d.). Informant Questionnaire on Cognitive Decline in the Elderly. National Centre for Epidemiology and Population Health. Retrieved October 13, 2026, from https://nceph.anu.edu.au/research/tools-resources/informant-questionnaire-cognitive-decline-elderly
Burton, J. K., Fearon, P., Noel-Storr, A. H., McShane, R., Stott, D. J., & Quinn, T. J. (2021). Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) for the detection of dementia within a secondary care setting. Cochrane Database of Systematic Reviews, 2021(7), Article CD010772. https://doi.org/10.1002/14651858.CD010772.pub3
Burton, J. K., Stott, D. J., McShane, R., Noel-Storr, A. H., Swann-Price, R. S., & Quinn, T. J. (2021). Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) for the early detection of dementia across a variety of healthcare settings. Cochrane Database of Systematic Reviews, 2021(7), Article CD011333. https://doi.org/10.1002/14651858.CD011333.pub3
Harwood, D. M. J., Hope, T., & Jacoby, R. (1997). Cognitive impairment in medical inpatients. I: Screening for dementia – Is history better than mental state? Age and Ageing, 26(1), 31–35. https://doi.org/10.1093/ageing/26.1.31
Jorm, A. F. (n.d.). Short form of the Informant Questionnaire on Cognitive Decline in the Elderly (Short IQCODE) [Measurement instrument]. Centre for Mental Health Research, The Australian National University. https://nceph.anu.edu.au/sites/prod.nceph.sca-lws06.anu.edu.au/files/English_short.pdf
Jorm, A. F. (1994). A short form of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): Development and cross-validation. Psychological Medicine, 24(1), 145–153. https://doi.org/10.1017/S003329170002691X
Jorm, A. F. (2004). The Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): A review. International Psychogeriatrics, 16(3), 275–293. https://doi.org/10.1017/S1041610204000390
Jorm, A. F., Broe, G. A., Creasey, H., Sulway, M. R., Dent, O., Fairley, M. J., Kos, S. C., & Tennant, C. (1996). Further data on the validity of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE). International Journal of Geriatric Psychiatry, 11(2), 131–139. https://doi.org/10.1002/(SICI)1099-1166(199602)11:2<131::AID-GPS294>3.0.CO;2-5
Jorm, A. F., & Jacomb, P. A. (1989). The Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): Socio-demographic correlates, reliability, validity and some norms. Psychological Medicine, 19(4), 1015–1022. https://doi.org/10.1017/S0033291700005742
Jorm, A. F., Scott, R., Cullen, J. S., & MacKinnon, A. J. (1991). Performance of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) as a screening test for dementia. Psychological Medicine, 21(3), 785–790. https://doi.org/10.1017/S0033291700022418
Quinn, T. J., Fearon, P., Noel-Storr, A. H., Young, C., McShane, R., & Stott, D. J. (2021). Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) for the detection of dementia within community dwelling populations. Cochrane Database of Systematic Reviews, 2021(7), Article CD010079. https://doi.org/10.1002/14651858.CD010079.pub3