Our study examined the performance of three different mood disorder screening scales for detection of major depressive disorder in older patients with cancer. To our knowledge, very few studies have examined depression and its detection in this population. The advanced average age of our population (81 years), the presence of severe comorbidities and the heterogeneity in cancer diagnoses as well as treatment plans ensure the representativeness of our population with regards to the general geriatric cancer population. Notable similarities exist between our population and those of large-scale prospective studies carried out in geriatric oncology [29–31]. Indeed, the socio-demographic characteristics of the population in our study, such as average age, living place and marital status are comparable to those found in studies by Kenis et al [29], Soubeyran et al. [30]and in the ELCAPA study [31]. Moreover, it is worth highlighting the fact that the functional status for basic daily living activities among our population is similar to that of the ELCAPA cohort (ADL ≤ 5: 29% as against31.5%) [31]. The general health status of the two populations was also comparable (PS ≥ 2: 52.7% as against49.9%)[31] in contrast to that in studies by Kenis et al. (PS ≥ 2: 29.6%)[29]and by Soubeyran et al. (PS ≥ 2: 22.8%) [30]. This discrepancy may be explained by the non-inclusion of exclusive palliative care patients.
Our study shows a statistically significant association between the presence of a MDD according to the DSM-V diagnostic criteria and a pathological screening using the GDS–15, the HADS-D and the DT. These results contradict the work of Rhondali et al., who found a moderate statistical association between the DSM diagnostic criteria for MDD and the 30-item version of the GDS, but none with either the HADS or the DT [20]. This discrepancy could be explained by methodological differences in terms of sampling size and population characteristics. In addition, in their study, Rhondali et al. used a global score for HADS (combination of depression and anxiety subparts). In our study, we only used the subpart depression (HADS-D), more specific to the symptoms of depression.
GDS–15 is currently the most widely-used screening tool for detecting depression in older patients. According to a meta-analysis carried out by Wancata et al. [32] it is 80% sensitive and 75% specific. In 2017, Saracino et al. [33] looked at the effectiveness of three scales for depression screening in patients with cancer over the age of 70. The authors found a sensitivity of 67%, a specificity of 88%, a ROC curve AUC of 0.88 (IC95%:0.80–0.95) and a LR+ of 5.51 for the GDS–15 [33]. In our study, ROC curve AUC was 0.68 (IC95%:0.45–0.91) and LR+ was 0.11 for this scale.
HADS is a scale designed to exclude any items relating to somatic aspects [27]. Among the general population it is reportedly 50% sensitive and 97% specific [27]. In two studies carried out among older patients with cancer, Rhondali et al. [20]and Saracino et al. [33]found for the HADS-D respectively sensitivities of 50% and 17% and specificities of 67% and 93%. In Saracino’s study [33], the ROC curve AUC of the HADS-D was 0.88 (IC95%:0.81–0.97) and the LR+ was 2.2632. In our study, we found a ROC curve AUC of 0.76 (IC95%:0.60–0.92) and a LR+ of 4.28 for this scale.
These results show an important heterogeneity regarding to the properties of the two scales. The absence of consensus regarding clinical cut-off for diagnosis may partly explain these discrepancies. Furthermore, the statistical performance of these two self-administered questionnaires appears disappointing. Indeed, a sensitivity of at least 80% and a specificity of at least 70% are considered necessary for depression screening in geriatric oncology [33]. This lack of sensitivity could expose older patients with cancer to a significant risk of under-diagnosis which could lead to an increase in the risk of suicide and morbidity/mortality. On the other hand, the lack of specificity could expose patients to over-diagnosis and a consequent risk of emotional breakdown and inappropriate medication prescriptions [33].
Beyond these statistical considerations, some of the questions included in these screening scales may seem inappropriate for older patients with cancer and could be misinterpreted in the context of a recent cancer diagnosis. For examples “Do you feel that your situation is hopeless?” or “Do you feel full of energy?” in the GDS–15 [25–26] or the statement “I get a sort of frightened feeling as if something awful is about to happen” in the HADS-D [27]. Such statements could draw the newly diagnosed cancer patient to focus on potentially negative experiences to come, as well as exacerbate anxiety or depression.
In our study, analysis of ROC curves and positive likelihood ratios underlined that the HADS-D was the most effective screening tool for detecting a MDD.
DT was designed for quick identification of individuals at risk of mood disorders (the time required for the test is less than one minute) [34]. Its use in patients with cancer is currently recommended by The National Comprehensive Cancer Network (NCCN) [35]. In our study, we experienced the original DT in a French older population and not the French Psychological Distress Scale version validated by Dolbeault et al36] whose cut-off was 3. Indeed, in elderly people, the original visual device seems to be more comprehensive than a 10 cm vertical line. In the same way, numerical scale is not appropriate for pain evaluation in this population. Our statistical analysis regarding the DT points to its potential usefulness in ambulatory medicine. From a statistical point of view, with a ROC curve AUC of 0.85 (IC95%:0.66–1) and a LR+ of 9.125, the DT carried out at 3 weeks by the primary care physician was more effective than the HADS-D. Analysis of the ROC curve and the likelihood ratio also demonstrated the significant predictive characteristics of the DT for detecting a MDD according to the DSM-V diagnostic criteria. Moreover, the strong positive correlation found between the DT score obtained during the initial assessment and during the reassessment by the primary care physician is good evidence that this test is highly reproducible. The superior performance of the DT for ambulatory patients may be explained by the fact that these patients had greater psychological resources available when they were reassessed by their primary care physician than they had in the initial hospital environment. Further studies are needed to confirm the performance of the DT in the ambulatory setting.
Studies carried out in older patients with cancer have shown an association between depression and various factors such as feminine gender, social isolation, grief, dependency, history of depression or metastatic cancer [37–40]. In contrast to these studies, we did not find any statistically significant association between the medical, psychological, socio-environmental factors studied and the detection of a MDD or abnormal mood disorders screening. This discrepancy can be explained by the heterogeneity of our population and small sub-group sample sizes.
Patients who presented a psychiatric history and/or were undergoing a psychotropic treatment were not excluded from this screening study. This was done purposefully to evaluate if mood disorders persisted despite ongoing treatment and to better characterize the depression and/or anxiety related clinical symptoms. Among the 10 patients who presented a confirmed MDD, only three had a known history of depression and were treated using antidepressants before the study. Given the small patient sample involved, we can assume that the inclusion of these patients had only a slight impact on the results obtained. However, these results raise concerns as to the efficacy of drug treatment. For patients recently informed of their diagnosis or of relapse of a cancerous pathology, it would have been interesting to study the types of treatment prescribed, dosing regimen and length of treatment. As expected, most cases of depression were not properly recognized. The use of screening scales led to the diagnosis of seven new cases of MDD. Furthermore, it is interesting to highlight the low acceptance rate of recommended psychological consultations (23%) as this is a critical issue for this patient population.
Our study design carries several limitations, including the limited number of patients reassessed, the small sample sizes for sub-group analysis and the non-inclusion of individuals presenting neurocognitive disorders. The methodology of our study was based on reassessment of mood disorders at 3 weeks by a different, impartial physician in order to detect the persistence of these disorders (as required to diagnose a MDD). However, reassessment was carried out on a voluntary basis by the physician and this might explain the significant number of losses to follow up. This design was chosen to preserve the observational nature of the study without interfering with care and to facilitate inclusion as well.
Future research, on a larger population, is needed to confirm our results with a view of measuring the sensitivity and specificity of these three mood disorder scales in older patients with cancer and to validate their use within this specific population. As we have argued elsewhere, use of DT for ambulatory patients may be a promising measure to recognize mood disorder in older patients with cancer.
Beyond the specific use of these screening tools, it seems of critical importance to call for depression in older patients with cancer to be recognized using a global, multidimensional and multidisciplinary approach, in which psychologists also play a role. Moreover, the fact that a majority of patients refuse the proposed psychological consultation should question us about our non-drug management methods in this particular population.