Introducing a translated cognitive assessment tool to general practitioners in Malta : a feasibility study

Abstract

The aim of the study was to translate and adapt the General Practitioner Assessment of Cognition (GPCOG) into the Maltese language and to validate the translated tool in older individuals within the local population. A framework for translation and cultural adaptation process was followed to create the Maltese GPCOG. A symmetrical translation was used for the Maltese GPCOG and the same format and construct was kept across items. A pilot study followed to ascertain that the tool was free from any language or cultural bias and was relevant to the Maltese culture and its intended use.

The targeted populations were recruited from within the community through a purposive sampling procedure. The tool was implemented with 207 individuals aged ≥55 years, including 54 individuals previously diagnosed with dementia at the local geriatric rehabilitation outpatient facility, (Karen Grech Rehabilitation Hospital) and Dementia Day Centre (St. Vincent de Paul Long Term Care Facility), as well as 153 healthy individuals who visited the 3 main health centres in the northern, central and southern regions. The majority consisted of females 66.8% (n=139) and 32.7% (n=68) were males, mean age was 68.2 years (r_55 – 95). Mean years of education of the sample was 9.1 ± 3.0 (r_0 – 19). Data collection took place over fourteen weeks and all assessments were administered by the researcher. The study was approved by the University Research and Ethics Committee (UREC) and the members of the Ethics Board Committee at KGRH.

Results from the translation analysis, test-retest and inter-rater analysis showed very good reliability statistics. Compared with the Maltese MMSE, the Maltese GPCOG Step 1 demonstrated excellent content validity. Correlation between the two tests was very strong (r = 0.932, p = < 0.001), which is stronger than that reported in the original GPCOG Step 1 study (r = 0.683, p = 0.001). Besides, the Maltese GPCOG takes about half the time to administer than the Maltese MMSE, making it more time-efficient and thus more practical in a busy general practice setting. The sensitivity and specificity of the Maltese GPCOG Step 1 (patient section), with a cut- off score of 7 points, was 92% and 97% respectively in this sample, the tool showed a very high predictive validity in identifying persons with cognitive impairment (area under the ROC curve = 95%). The Maltese GPCOG Step 2 (informant interview), collects collateral information about the patient from a reliable informant to establish any observed cognitive or functional changes over the previous 5-to-10 years. Step 2 was administered with 25 individuals who scored 5, 6, 7 or 8 on Step 1, as per instructions of the tool.

The Logistic regression model on Step 1 indicated that a score 0-to-6 points predicted cognitive impairment and a score of 7-to-9 points predicted normal cognition. The diagnostic accuracy of the Maltese GPCOG Step 1, was confirmed by the likelihood ratio test, which indicated that for every 1 unit increase on Step 1, the probability of cognitive impairment decreases by 88.5%.

In conclusion, the Maltese GPCOG is very brief and simple. The results of the present study confirm that the Maltese version is both reliable and valid in the assessment of cognitive impairment in the local population. In addition, the brief duration (3.5 minutes ± 1.49) make it an efficient instrument suitable for cognitive screening at primary care level in Malta.