Prevalence of Delirium in Older Medical Inpatients in Tanzania

To the Editor: Delirium is an important and common disorder in older medical inpatients in high-income countries.1 The proportion of medical inpatients with delirium at hospital admission is reported to be between 10% and 30%.2 Previous high quality research describing the prevalence of delirium in sub-Saharan Africa (SSA) is scarce, and estimates vary widely.3 The only published study of older medical inpatients from SSA is a Nigerian study reporting a delirium prevalence of only 9.4% in a sample of 106 individuals.4 Adults aged 60 and older sequentially admitted to the acute medical wards of Kilimanjaro Christian Medical Centre, Northern Tanzania, one of four tertiary referral hospitals in Tanzania, with approximately 800 inpatient beds, were recruited and screened on the day after admission. Individuals who died or were discharged before consent was obtained were excluded. The prevalence of delirium at this assessment was taken as the point-prevalence on admission. Data were collected from January 14 to February 3, 2015, and from March 6 to July 10, 2015. Trained medical doctors (EGL, SMP, ATD, LT) used the Confusion Assessment Method (CAM) in conjunction with a neurocognitive assessment; a positive CAM result was taken as a diagnosis of delirium. An informant history, which is vital for identifying the acute and fluctuating course of delirium and distinguishing between delirium and dementia, was also obtained. The neurocognitive assessment comprised a number of bedside tests designed to assess various cognitive domains such as attention and concentration. Highly skilled local health professionals (AK, JC, CL), who are fluent in English and Swahili and experienced in psychiatric research interviews, facilitated the assessments. The CAM has been widely used and validated in many countries and healthcare settings, including low- and middle-income countries such as Brazil.5 It has a pooled sensitivity of 82% (95% confidence interval (CI) = 69–91%) and specificity of 99% (95% CI = 87–100%),5 but to the knowledge of the authors of the current study, there is only one previous example of the CAM being used for research in SSA.6 Five hundred and ten admitted individuals provided consent and were assessed, and 99 (16.2%) were excluded. Of those included, 284 (55.6%) were male. The median age of those admitted was 75 (interquartile range 67–81, range 60–104). Most participants were reviewed the day after admission (83.5%), 10.2% 2 days after admission, 5.4% 3 days after admission, and the remaining 0.9% within 4–7 days of admission. Fifty-four individuals could not be assessed using the CAM because of lack of consciousness. The estimated point prevalence of delirium at admission was 19.5% (Table 1). Delirium was more common in men and older individuals. A delirium prevalence on admission was found of 19.5%, similar to that found in high-income countries (10–30%).2 A lower prevalence of delirium might have been expected based on previously published data.4 This is an important finding because the majority of adults aged 60 and older worldwide currently live in low- and middle-income countries, and this proportion is increasing rapidly.7 In SSA, the percentage of older adults as a proportion of the general population is expected to increase from 6% to 10% by 2050.7 This information could be of use to policy-makers when planning inpatient services for older adults across SSA. Delirium was associated with male sex. This pattern was found in a large study of older adults admitted to three SSA tertiary-care hospitals, and was attributed to an inequality in accessing hospital care according to sex.8 There may also be a higher disease burden among men aged 60 and older; for example, stroke incidence in Tanzania is higher in men,9 which may partially explain this finding. Previous work has suggested that delirium is underdiagnosed in at least two-thirds of cases.2, 10 Although healthcare resources are limited across much of SSA, and health priorities must be balanced, training clinical staff in the use of the CAM and other cognitive assessments may help to increase diagnosis rates. We would like to thank Gloria Temu, Nyasatu Chamba, John Kissima, Nuru Mwaluwinga, Editruda Gamassa, Victoria Ferguson, Gillian Tough, and the individuals and family members who took part for their contributions to this study. Conflict of Interest: None. Author Contributions: Lewis, Paddick, Gray, Walker, Dotchin, Urasa: study concept and design. Lewis, Paddick, Banks, Tucker, Duinmaijer, Kisoli, Cletus, Lissu, Urasa: acquisition of data. Lewis, Paddick, Banks, Gray: data analysis and interpretation. All authors: drafting and revising manuscript, approval of final version. Sponsor's Role: Partly funded by Grand Challenges Canada (Grant 0086–04).

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.