Screening for Depression in People with Kidney Failure

A large proportion of patients with CKD before and after initiation of maintenance dialysis suffer from major depressive disorder, such that about 20% of those with non-dialysis CKD and 25% of those with kidney failure are affected (1). More daunting is the actuality, validated in multiple observational studies, that the presence of either major depressive disorder or depressive symptoms ascertained by self-report questionnaires is independently and consistently associated with adverse kidney, cardiovascular, survival, and quality of life outcomes in these patient populations (2,3). Therefore, it becomes imperative to actively recognize the presence of major depressive disorder or depressive symptoms, such that the Centers for Medicare and Medicaid Services’ ESRD Quality Incentive Program has set forth requirements for routine depression screening for patients with kidney failure.

One major challenge that remains in the comprehensive clinical care of our patients is that they often express a burden of symptoms that have considerable overlap with more organic or somatic depressive symptoms, which makes differentiation from a psychiatric diagnosis of major depressive disorder difficult. These symptoms are especially experienced during advanced stages of kidney disease or may be exacerbated due to comorbid medical illnesses—such as diabetes mellitus, obstructive sleep apnea, and heart failure, to name just a few—and include fatigue, difficulty concentrating, sleep disturbances, loss of appetite, weight changes, and psychomotor agitation or retardation. Following this premise, a major limitation of self-report depression questionnaires is that they include items that assess these symptoms for a final total score, depicting the severity of symptoms. Consequently, patients with a high burden of symptoms related to kidney disease may have falsely inflated depressive symptoms by these measures. In addition, most questionnaires have validated cutoff scores that are used to ascertain risk of having a clinical diagnosis of major depressive disorder that would alert the clinician to pursue further diagnostic workup or refer the patient for mental health specialty management. However, these questionnaires and cutoffs were initially derived in the general population, and most of the data that previously existed regarding their psychometric behavior could only be applied to populations without chronic diseases such as CKD. Consequently, several investigators have conducted well-designed studies to validate these questionnaires in our patient populations so they can be more informatively used for both clinical and research purposes.

In the article titled “Depression Screening Tools for Patients with Kidney Failure: A Systematic Review” published in this issue of CJASN, Kondo et al. (4) conducted a comprehensive systemic review to identify depression screening tools and thresholds appropriate for patients with kidney failure on maintenance dialysis. The authors should be commended on undertaking this overdue task of filling an important knowledge gap in the multidisciplinary care of a very vulnerable population, namely confirming the usability of tools aimed to identify an independent risk factor associated with future adverse outcomes. Importantly, they selected studies comparing various screening tools to Diagnostic and Statistical Manual of Mental Disorders–based gold-standard diagnosis of major depressive disorder. They provide a thorough and easily understandable summary of the data on this timely, practical, and patient-centered clinical question.

The authors conclude that the Beck Depression Inventory II (BDI-II) is the best-studied tool, with a cutoff of ≥16 providing a balance of sensitivity and specificity for identifying depression in patients with kidney failure (4). This choice is largely informed by the scarcity of data about other measures, as the BDI-II has been by far the most investigated self-report questionnaire used and studied in patients with kidney failure; therefore, most studies available in the published literature for this systemic review used the BDI-II. The authors further justify their recommendation based on two studies demonstrating that the BDI-II was more accurate than the Cognitive Depression Index subset (which excludes somatic symptoms) for identifying individuals with major depressive disorder (4,5). They report that for the four studies evaluating a BDI-II cutoff of 16, sensitivity ranged from 0.81 to 0.92 and specificity from 0.58 to 0.87. Of the cutoffs presented for the BDI-II, a cutoff of 16 was one of few reported in more than two studies and, appropriate for a screening tool, favoring higher sensitivity over slightly lower but acceptable specificity (6,7).

However, some of the data presented argue against the use of the BDI-II as the optimal depression screening tool in this patient population. First, it may not be the most accurate of the currently studied questionnaires. The authors mention that in head-to-head comparison, the Depression Inventory–Maintenance Hemodialysis scale performed better than the BDI-II for identifying depression, albeit only in one study (4). The absence of multiple studies evaluating the other available scales should not necessarily count as a strike against their use in clinical practice if well-designed existing studies reported promising results, as the authors noted is the case for the Patient Health Questionnaire 9 (PHQ-9) or the Geriatric Depression Scale 15. There are also other available measures such as the self-reported Quick Inventory of Depressive Symptomatology (QIDS-SR₁₆) that have been validated in patients with stages 2–5 CKD (8) and used in clinical trials of depression treatments in patients with CKD (9) and kidney failure (10), although studies of such tools did not meet the criteria for inclusion in this systematic review. The authors also note there were no studies of the PHQ-2, despite its common use as a screening tool in clinical practice. Second, the BDI-II may generate more false positives than other scales because somatic symptoms, precisely those commonly expressed in patients with kidney failure, are overemphasized because equivalent weights are not given to each of the nine core symptom domains for major depressive disorder. Other measures such as the QIDS-SR₁₆ or PHQ-9 that give equal weight to the various domains in calculating the final score may be less likely to overestimate depression prevalence and severity in patients with kidney disease.

Furthermore, there are multiple practical considerations to take into account when considering the best screening tool to identify depression in patients with kidney failure on maintenance dialysis. First, the BDI-II has 21 items, more than other self-report questionnaires (e.g., the QIDS-SR₁₆ has 16 items, and the PHQ-9 has nine items), and thus is longer and more time consuming to administer to patients who already have a high burden of health care contact. Second, as noted by the authors, use of the BDI-II requires a fee, whereas other scales such as the QIDS-SR and PHQ-9 are publicly available free of charge for use in clinical practice. Overall, when considering its limitations, we would argue that the BDI-II is likely not the most practical or accessible tool to use in clinical practice, and favor lesser-studied but shorter, freely available measures such as the QIDS-SR₁₆ or the PHQ-9.

Due to the nature of the available data, this systematic review focused primarily on individuals receiving maintenance hemodialysis. Unfortunately, only four of the studies included patients receiving peritoneal dialysis, and none were conducted exclusively in such patients, highlighting knowledge gaps among this population. Also, the validity of self-report questionnaire in nondialysis CKD patients was not undertaken, which becomes imperative as CKD comprises a much larger segment of our patient population, and identifying risk in patients with earlier stages of disease would potentially lead to interventions that may prevent adverse outcomes. Understandably, this was outside the scope of this work.

One important consideration about depression screening is worth noting. Self-report questionnaires, whether validated in populations with kidney disease, lack criteria for the inclusion of either sadness or anhedonia (loss of interest or pleasure) to be part of the total score that defines the cutoff for the presence of major depressive disorder. Therefore, a confirmatory interview by a health care provider or social worker is crucial to make a major depressive disorder diagnosis, which requires that one of these symptoms (sadness or anhedonia) be present as one of the five or more symptom domains, including sadness, anhedonia, worthlessness or inappropriate guilt, recurring thoughts of death or suicide, appetite and/or weight disturbance, sleep disturbance, psychomotor agitation or retardation, fatigue, and difficulty concentrating. In addition, identifying patients with active suicidal ideation is crucial given that those patients will require further evaluation to determine whether urgent medical attention and management are required. For nephrology providers practicing in the clinical setting, it is often most pragmatic to identify patients with depressive symptoms first by administration of self-report questionnaires to identify risk if they meet the validated cutoffs, followed by confirmation of (1) the presence of sadness or anhedonia; (2) the presence of symptoms for at least 2 weeks for major depressive disorder diagnosis; and (3) the absence of active suicidal ideation, which may require urgent attention or vigilant follow-up.

Screening for depression in patients with kidney failure on maintenance dialysis has become a greater priority in clinical care. Kondo et al. present a comprehensive systematic review detailing the sensitivity and specificity of various self-report screening tools in patients with kidney failure. However, questions remain about the most accurate tools to identify patients at risk, particularly given the absence of extensive evidence for multiple promising, brief, publicly available questionnaires that may be more practical for use in clinical practice.

Disclosures

L.P. Gregg reports employment at Baylor College of Medicine, Houston, Texas, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas and Veterans Affairs Health Services Research and Development Center for Innovations in Quality, Effectiveness and Safety, Houston, Texas. S.S. Hedayati reports receiving honoraria from the American College of Physicians for participation in Nephrology Medical Knowledge Self-Assessment Program and honoraria from the American Society of Nephrology Post-Graduate Education Program, and has served as a scientific advisor or member of the study sections at the American Heart Association, Medical Knowledge Self-Assessment Program Nephrology Committee at the American College of Physicians, and the American Society of Nephrology In-Training Examination Committee.

Funding

S.S. Hedayati was supported by the National Institute of Diabetes and Digestive and Kidney Diseases grant R01DK124379 and by the Yin Quan-Yuen Distinguished Professorship in Nephrology.