Gatos Questionnaire for Early Detection of Suspicious Signs for Alzheimer Disease and Related Dementia Syndromes - GQEDSS - ADRDS Q160 v 1.0. Preliminary Results
Gregory Tsoucalas*, Stamati Bourelia, Anastasios Markellos, Vasiliki Kalogirou, Styliani Giatsiou, Olga Repana, Presveia Gatou, Ifigeneia Georgousi, Anna Maria Xanthi, Andromachi Korenti, Nikolaos Kosmas, Antonios Antoniou, Konstantinos Gatos
Interdisciplinary Study Team, Neurological Clinic-Hellenic Reference for Alzheimer Disease and Dementia Related Syndromes Agios Georgios, Alykes Volou, Greece
To cite this article:
Gregory Tsoucalas, Stamati Bourelia, Anastasios Markellos, Vasiliki Kalogirou, Styliani Giatsiou, Olga Repana, Presveia Gatou, Ifigeneia Georgousi, Anna Maria Xanthi, Andromachi Korenti, Nikolaos Kosmas, Antonios Antoniou, Konstantinos Gatos. Gatos Questionnaire for Early Detection of Suspicious Signs for Alzheimer Disease and Related Dementia Syndromes - GQEDSS - ADRDS Q160 v 1.0. Preliminary Results. International Journal of Psychological and Brain Sciences. Vol. 1, No. 3, 2016, pp. 69-85. doi: 10.11648/j.ijpbs.20160103.15
Received: November 9, 2016; Accepted: December 9, 2016; Published: December 20, 2016
Abstract: Dementia presents a cluster of syndromes the effect human brain. It rises various social-economic and health care issues that are difficult to be confronted. Early detection of the possibility of an examinee to be inflicted by a dementia syndrome is an all neurologists quest. Our interdisciplinary team for some year know created a questionnaire that depicts a series of items that could have been changed before the diagnosis of a dementia to be set. Material and Method: More than 180 items were included in the questionnaire. The validation process excluded 26 items. Then n1=55 examinees enrolled to our study in order to be watched for 5 years (n2=26 were diagnosed with dementia). At the end of this period 26 were inflicted by a dementia (mostly Alzheimer disease and vascular dementia). All answered the GQEDSS - ADRDS Q160 v 1.0 (3 demographic items and 157 Likert scale items), and the MMSE questionnaires. A statistical analysis of the questionnaire was followed. The scree test formed 54 factors, followed by a key item analysis (15 key items) to simplify factors analysis' complexity. Convergent or Criterion validity of the GQEDSS - ADRDS questionnaire was determined by establishing its correlation to the MMSE score using the Pearson’s correlation coefficient. Discriminant validity tested whether concepts or measurements that are supposed to be unrelated are, in fact, unrelated. Internal consistency validity of the GQEDSS - ADRDS was determined by calculating Cronbach alpha coefficient. Test-retest reliability (stability) was also performed. Results: The Bartlett Test of Sphericity was 2527.2 and it was significant (p<0.0015). The Kaiser-Meyer-Olkin Measure of Sampling Adequacy was equal to 0.801 showing that the data is suitable for factor analysis. The 160 items were analysed via maximum likelihood extraction method using a Varimax rotation. The correlation coefficients presented p<0.0005. The cut-off points (total summary of 700 scoring points) were set to a: a) < 120, there is a relatively small change, b) 121-180, there is a mediocre change, c) 181-220, there is a relatively good change, and d) >221, there is a relatively high change. Discussion: Although our study presents an intriguing questionnaire, our sample does not meet the Stevens' criteria. Small sample, examiners' subjectivity, non-randomized sample - Convenience, or consecutive sampling (all clients in the outpatient clinic), questionnaire longevity could present a series of BIASes. More study, co-operation with other specialized centres, and a better methodology are required.
Keywords: Dementia, Early Detection, Questionnaire
Dementias present a cluster of clinical syndromes characterized of a widespread progressive deterioration of cognitive abilities. This impairment interferes with normal daily functioning posing considerable challenges to patients and their caregivers. Over 70 different causal entities for dementia have been described. Alzheimer's disease is the most common cause of dementia but not all dementia is due to Alzheimer's. Thus vascular dementia has a significant prevalence among the general population. Despite the heterogeneity in etiology, dementias affect the mental competence, memory and learning abilities, speech, orientation, balance, gait, perception, concentration, judgment, problem solving ability and social functioning. Emotions and personality are impaired, the patient progressively deteriorates and finally becomes incapable even to complete elementary mental functions. Dementias are a serious public health problem, with many social and economic dimensions, as care needs for dementia patients build up over time [1-2].
It is proven that dementias are not restricted to elderly only, but they could also attack people younger than 60 years old. The progressive and mostly irreversible nature of dementias along with their impact on health system and society calls for early diagnosis. Early detection of dementia is essential to guide front-line health care practitioners in further clinical evaluations and give neurologists time to develop and implement of preventive and therapeutic treatments .
International medical literature provides a series of shorter or longer screening tools, batteries combining small cognitive tests and questions to unveil possible brain fuctional damage. The most important are: Seven minute neurocognitive screening battery (7MS), the Memory Impairment Screen (MIS), the clock drawing test, the cube-copying test, the DemTect, the AB Cognitive Screen (ABCS), the AD8, the Montreal Cognitive Assessment (MoCA), the Short Cognitive Performance Test, the Addenbrooke’s Cognitive Examination Revised (ACE-R), the Memory Alteration Test (MAT), the Memory Orientation Screening Test (MOST), the Mini Mental State Examination severe (MMSE-s), the Severe Impairment Battery (SIB), and finally the first version of the worldwide famous Mini Mental State Examination (MMSE) [4-5].
The MMSE, as a brief mental status examination is a useful short screening tool, which must be accompanied by a detailed assessment of individual cognitive and should be always subject to criticism when it is not aligned with the patient’s clinical picture. It has been emerged as a useful tool for physicians regardless of specialty, and allows a gross approach to the patient's condition, irrelevant of medication or comorbidities. The MMSE ranges are 30 scoring points for no, 26-29 for questionable, 21-25 for mild, 11-20 for moderate, and 0-10 for severe dementia. However, MMSE does not interfere with quality of life, it is not a prognostic tool for dementia course, and it does not discriminate reversible disease, while other conditions, such as delirium, bipolar disorder, or mental retardation often lead to low scores in MMSE and patients misclassification [6-7].
This study focuses on the earliest changes in neuropsychological functioning, hoping to present a new useful screening questionnaire for the early detection of impairment by reviewing a plethora of cognitive, functional and socializing data. Our questionnaire is separated into 4 sectors: i) demographic data, ii) concerning the change in upper cerebral functions, iii) concerning the change in cerebral function in relation to the interaction between both the external stimuli and the environment, and iv) concerning the changes in relation to oneself, while it contains 160 points-characteristics to be evaluated.
2. Material and Method
A panel of ten, dementia specialized health professionals were engaged in the validation process. For every item to win its place in the questionnaire, every team member assessed the item as "necessary", "useful but not necessary" and "not necessary". Calculating the content validity ratio, all questions scored between 0.5 and 1 were included in the questionnaire (26 more items were excluded due to minor scoring). The final form of the GQEDSS - ADRDS comprises firstly of a part consisted of 3 domains of general demographic information related to person in examination and secondly the main part consisted of 157 items (questions). This second part questions are rated on a Likert scale. These 160 items form the GQEDSS - ADRDS scale. The questionnaire is presented in the appendix section of the paper. The 160 items total score is used to evaluate person's probability to experience a dementia syndrome in the future. For the individualized and holistic clinical assessment of the patient, however, both parts, should be considered.
Fifty five persons (n1=55) visiting our outpatient sector were enrolled in the study. The only criterion was an MMSE score between 0 and 30. Thus relatively "brain healthy" individuals. Informed consent brochures were administered and the study was fully conformed with Helsinki declaration principles and Greek Medical ethics law (Law 3418/2005). To demonstrate the discriminant validity of the GQEDSS - ADRDS, the MMSE was also applied. To fill in both questionnaires, the patient, the relatives, the caregivers and the specialized personnel cooperation was in some cases necessary. Both were filled at the same day for every patient, while for the summary of the fifty five patients to be enrolled in the study, a time period of five months (January - May 2011) was necessary. For the test-retest validation a period of 48 hours was used for 12 participants to answer both scales once more. All participants have been under then under surveillance for the change of a dementia syndrome to inflict them for further five years (until January 2016). From the total number of fifty five participants, twenty six (n2=26) presented dementia. Then, the sample n2 was used for the significance of various key items to be measured.
All items were coded and scored and the completed questionnaires were included in the data set. Individual unanswered items were excluded from the analysis. All the data were entered, checked for missing values and analysed using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA) and SAS version 7.0 (SAS Institute, Cary, NC, USA) statistical programs. Descriptive statistics, such as the mean, standard deviation and skewness were used to describe the main variables. However, due to small study sample, parametric methods were used to further analyze the data. Statistical significance was set at p<0,05. Data distribution was tested with Kolmogorov-Smirnov test και P-P plots.
The validity of a questionnaire is the extent to which the questionnaire measures the study concept or, otherwise, the variate that it is supposed to measure. Increase in validity of a questionnaire is associated with a decrease in systematic error. However due to the small sample the evaluation of the validity of the GQEDSS - ADRDS presents a priori questionable results. The questionnaire validation included factor analysis, convergent or criterion validity, subscale validity, discriminant validity, key items analysis, and ROC analysis.
Exploratory factor analysis (EFA) was conducted to identify a viable factor structure. EFA, using principal component extraction method with Varimax rotation, was conducted for all participants to determine the factor structure of the 160 items of the GQEDSS - ADRDS questionnaire. The sample size of 26 patients did not satisfied the criteria set by Stevens  (factors loadings require n ≥300). Factor analyses were repeated until a solution was attained in which all items included in the analysis met all Stevens' criteria [9, 10]. The scree test used to determine the number of factors to retain and rotate. From the analysis 54 factors appeared to be valid. Convergent or Criterion validity of the GQEDSS - ADRDS questionnaire was determined by establishing its correlation to the MMSE score using the Pearson’s correlation coefficient. Subscale validity was evaluated by examining the subscale correlations. Discriminant validity tested whether concepts or measurements that are supposed to be unrelated are, in fact, unrelated. A successful evaluation of discriminant validity shows that our questionnaire measuring mental status is not highly correlated with other questionnaire designed to measure theoretically different concepts. The discriminant validity examined using a sample of twelve random persons (n3=12). Key items evaluation was conducted for the factors' complexity and variety to be simplified.
A receiver operating curve (ROC) analysis was conducted to obtain the cut-off level of GQEDSS - ADRDS total score for differentiation between subgroups of patients formed on the basis of their disease probability, calculating the respective areas under the curve (AUC). The areas under the ROC curve (AUC) with standard error and 95%CI were calculated using the maximum likelihood estimation method, which has the advantage of being free of assumption about the Gaussian distribution of underlying variables. Furthermore, the sensitivity and specificity of different cut-off points of GQEDSS - ADRDS total were estimated using the MMSE score (28 vs 29 or 30) as a gold standard method.
The reliability or precision of a questionnaire concerns the degree of stability or consistency with which the questionnaire measures the concept that it is supposed to measure. The estimation of reliability of the GQEDSS - ADRDS questionnaire included internal consistency reliability, test-retest reliability and inter-observer reliability. Internal consistency validity of the GQEDSS - ADRDS was determined by calculating Cronbach alpha coefficient. A Cronbach α coefficient value of 0.7 indicates sufficient reliability for research purposes and suggests that items are interdependent and homogeneous in terms of the construct they measure. For clinical applications a > 0.8 is desirable [11-13]. Test-retest reliability (stability) (n3 patients) indicates the stability of patients’ response in time and it was determined by calculating ICC (intraclass correlation coefficient: the error in measurements as a proportion of the total variance) between the total scores of the initial assessment of the GQEDSS - ADRDS and the total scores of the reassessment after 3 days. Because this coefficient does not correct for systematic differences and agreement by chance, the scores of the 2 assessments were tested for systematic differences by using the Paired t-test. Inter-observer reliability is a measure of reliability used to assess the degree to which different judges or raters agree in their assessment decisions. Inter-rater reliability is useful because human observers will not necessarily interpret answers the same way; raters may disagree as to how well certain responses or material demonstrate knowledge of the construct or skill being assessed .The Inter-observer reliability was examined using the sample of n3 random patients.
Demographics are shown in [Table 1]. The Bartlett Test of Sphericity was 2527.2 and it was significant (p<0.0015). The Kaiser-Meyer-Olkin Measure of Sampling Adequacy was equal to 0.801 showing that the data is suitable for factor analysis. The 160 items were analysed via maximum likelihood extraction method using a Varimax rotation. Fifty four factors with eigenvalues of over 1 were identified. We used the scree test to determine the number of factors to retain and rotate, which suggested a 54 factor solution. Then a key items evaluation was conducted, naming 15 key items (dark coloured background in the questionnaire) as most important after a horizontal statistic analysis (mean value of n2 items) [Table 2]. The GQEDSS - ADRDS 15 key items and total score with MMSE score exhibited excellent correlation. The correlation coefficients presented p<0.0005.The above result indicated very strong correlation between the 2 questionnaires and satisfied the criterion validity. The area under the curve (AUC) ofGQEDSS - ADRDS total score was 0.779 (89%CI 0.82- 0.90 p<0.015), with cut-off points being set as: a) < 120, there is a relatively small change, b) 121-180, there is a mediocre change, c) 181-220, there is a relatively good change, and d) >>221, there is a relatively high change. The internal consistency of the 160 items of the GQEDSS - ADRDS was measured with Cronbach's alpha and yielded a value of 0.81, which indicates excellent internal consistency. All values are higher than 0.62 suggesting that items are interdependent and homogeneous in terms of the construct they measure. The paired samples t-test between initial assesment and reassesment of questionnaire indicated no statistically significant differences. ICC between initial assesment and reassesment of the test ranged between 0.862-0.959 (p<0.0005). The above results of stability indicated that GQEDSS - ADRDS total scores were remarkably consistent between the two occasions, proving test-retest reliability. The scores of GQEDSS - ADRDS questionnaire from the two different specialised doctors were highly correlated (ICC=0.863 to 0.959, p<0.0005) which proved that the raters agreed as to how well certain responses demonstrate knowledge of the construct or skill being assessed, proving inter-observer reliability.
|Sex||N||Living area||Education. In years|
|Urban||Rural||Island||DS in 5 years||0 1-9 >10|
|Male (n)||34||9||22||3||15||2 7 6|
|Female (n)||22||10||11||1||11||7 3 1|
|Total (n)||n1: 56||19||33||4||n2: 26||9 10 7|
|Sex||Age. Distribution||Farmers – Blue. Colars||White. Collars||Scietists - Business men||House keeping|
Blue collars: In Greece farmers and peasants should be included.
House keeping: In Greece house keeping is considered a profession, without being paid as such though.
DS: Dementia Syndrome.
|Number||Number inside GQEDSS - ADRDS||Item|
|3||Emotion - emotional tension||No 15|
|4||Thought disruption – concreteness of thinking||No 19|
|5||Memorize - integrate||No 24|
|6||Stress disorders||No 58|
|7||Ability in fine - combinatory movements||No 59|
|8||Dealing with daily problems / activities||No 69|
|9||Conversation flow - subject flow||No 78|
|10||Action - reaction time, response time||No 83|
|11||Balance during moving or walking||No 107|
|12||Weakness in orientation||No 110|
|13||Isolation - pullback||No 122|
|14||Personal hygiene||No 130|
|15||Disorder in sleeping pattern – insomnia, reversal, excessive sleep||No 153|
An assessment questionnaire in medicine transform structured information into numerical values  and so does the GQEDSS - ADRDS. Our purpose was not to replace simpler tools such as MMSE or most modern and sophisticated as the SIB, or the MMSE-s (Mini Mental State Examination-severe) , the TSI (Test Severe Impairment) , the SCIP (Severe Cognitive Impairment Profile) , the NTG-EDSD Screening Instrument by the American Academy of Developmental Medicine  and the Addenbrooke's Cognitive Examination-Revised for the Greek population (ACE-R) , but to fill the qualitative gaps in the assessment of the probability of a person to be inflicted by a dementia syndrome in the future. Our questionnaire could also be used as an alternative, or in combination with the validated tools already in use. The GQEDSS - ADRDS contributes to a different approach, introducing a communication pathway between physicians and clients, assessing some new areas that could be early impaired. It provides information that medical personnel, relatives and caregivers can take advantage of, in order to achieve a better scheduled confrontation against the possibility of an infliction by dementia syndromes (cognitive exercises, puzzle games, chess, Sudoku, riddles, socializing, medication etc). However, the validity of the new scale was based on a correlation with MMSE scores 28, 29 and 30, which when standing as unique measurements are generally considered unreliable. In addition, MMSE measures cognition whereas GQEDSS - ADRDS scale measures cognitive, functional, physical, social and personal performance. As MMSE is a specific and widely available scale, we proceeded with the comparison between these two scales, despite the aforementioned differences. A disadvantage of GQEDSS - ADRDS is that its completion time is estimates between 60 and 75 minutes, depending on the person being examined (age, education, etc).
The new GQEDSS - ADRDS scale, is a combination of an extensive experience of about a 30 years battle with demented patients , and items borrowed by predicting dementia batteries that are availliable for use [19-20,22-25]. However, GQEDSS - ADRDS could present a series of statistical biases. The sample was small in comparison with the number required by the Stevens criteria. All participants were visitors of the outpatient infirmary, thus somehow all felt some changes, suspicious for mental diseases. The MMSE scale which was used for the validation of the GQEDSS - ADRDS questionnaire is in a way one-dimensional tool (mainly memory impairment scale). Test-retest and the 5 year follow up by different members of the interdisciplinary team poses the risk for subjectivity. Extensive experience of the team's members could in some cases magnify an item that has minor significance during the interview-completion of the questionnaire. The sample was non-randomized - convenience, and/or consecutive sampling (all clients in the outpatient clinic). It is clear that further study is needed, much bigger sample, possibly through a multi-centred effort. Further key items validation in a try for the new tool to be shortened is somehow required, as its complexity and longevity make it difficult to be managed, and rather inapplicable in the Primary Health Care System. New methodology in agreement with Stevens criteria is also necessary. Our tool, depicts though an interdisciplinary team's of experts in dementia first attempt to predict some potential risk for the participants. This attempt could be further ameliorated in the near future.
The GQEDSS - ADRDS scale, as a new tool, was validated and seems to offer new items for the early discovery of signs that could possibly measure the risk for an individual to experience a dementia syndrome in the future. More work is needed, thorough study of a bigger sample, new modalities to be examined, new experts to participate in further inquiries, as the primary material in the new tool is at least intriguing and promising.
Appendix: The GQEDSS - ADRDS Questionnaire v 1.0