Members of PatientsLikeMe provide demographic information such as age, race, educational status, health insurance coverage, and location to help match demographically similar patients. They also identify primary and comorbid medical conditions from among more than 2500 conditions in the system or enter a new condition. Patients can complete a list of their symptoms and rate the current severity of symptoms (none/mild/moderate/severe), both of which can be updated over time. In addition, patients are invited to rate the effectiveness of treatments, including prescription and over-the-counter medications. These treatment evaluations also include treatment burden, adverse effects, adherence, and cost. Furthermore, the PatientsLikeMe platform database collects free-text entries, in which members provide qualitative, narrative information. Individuals were not required to complete all fields in the PatientsLikeMe system for their data to be included in this database analysis.

In the retrospective database analysis, data were examined from the entire PatientsLikeMe community of patients who reported having MDD (“the total MDD community”). This community comprised 17,166 individuals as of June 1, 2015. Patients who selected MDD among their conditions in the PatientsLikeMe system were prompted to provide feedback on specific symptoms, including depressed mood, fatigue, insomnia, inability to experience pleasure (anhedonia), low self-esteem, lack of motivation, and problems with concentration.

The custom survey was developed by all authors using information obtained during the study 1 analysis and was written in a way that was understandable to individuals with an average Flesch–Kincaid [21] reading grade of 7.5. The survey (Multimedia Appendix 1) included items comprising demographic information and patient characteristics (six items); medication history (three items); patient experience of depression in brief (three items); the patient health questionnaire (PHQ-8) depression scale (eight items), which is an adaptation of the clinically used PHQ-9 scale [22] that omits one item relating to suicidal thoughts and is a commonly used and accepted brief screening measure of depression [23]; patient prioritization of depressive symptoms (seven items); patient experience of memory, thinking, and concentration (seven items); patient-reported cognitive function using an adapted version of the perceived deficits questionnaire (PDQ-5), which is a cognitive function severity scale ranging from 0 to 20 (0=low perceived impairment) [24] and has been validated in patients with major depression [25], and using patient perception of cognitive symptoms and depressive relapse (11 items in total); and seeking advice and life adjustments (four items).

Longstanding (n=288; rated depressed mood at least six times during membership and active within the last 30 months) and recently active (n=3112; active within the last 90 days) members of the “total MDD community” were individually invited to participate in the PIVOT survey (total of 3400). Members were invited to participate via private message through the PatientsLikeMe platform. The private message contained research patient information and links to accept or decline the invitation to participate in the study.

Patients who accepted the invitation were directed to the electronic survey. To maximize participation, an email reminder was sent to users who did not accept or decline the survey invitation within 3 days. An email reminder was also sent to users who partially completed the survey within 3 days of their last survey data entry. The patient flow is shown in Figure 1. Only those who completed at least half of the survey were included in the analysis sample (N=525). Patients were not compensated for their participation.

Similar to study 1, the total frequency of survey respondents (n) was used, and missing data were not imputed. Data analysis of the prospective survey questionnaire was performed in line with that reported for the retrospective database analysis in study 1. Response rates were calculated using the checklist for reporting results of the internet e-survey criteria [26]. Severity levels on the PHQ-8 depression scale were scored according to Centers for Disease Control and Prevention guidelines (none = 0-4; mild = 5-9; moderate = 10-14; moderately severe = 15-19; and severe = 20-24) [23]. If data were missing for any question, the PHQ-8 score was deemed to be missing. The Spearman rank correlation was used to measure the degree of association between cognitive symptoms and depressive severity. Some demographic information that was not directly requested in the survey was taken from the PatientsLikeMe platform database and included the country of origin and comorbidities. Survey data were reported in cases where both platform and survey data were available.

Open-response questions were reviewed and analyzed for themes. Commonly reported themes were assigned a unique code, and the frequencies of codes were counted. Participants responding with multiple themes per question were counted for each theme reported; thus, the themes were not mutually exclusive by participant and question.

The “total MDD community” comprised 17,166 individuals. The majority were women (77%); were white (89%); lived in the United States (75%); and had at least some college education, an undergraduate degree, or a professional degree (80%; Table 1). The MDD characteristics of this population are shown in Table 1. Overall, 78% of individuals reported that they had, at some point, received treatment for MDD. In addition, 65% indicated the current severity of their depressed mood as moderate/severe (compared with 25% reporting mild and 10% reporting no depressed mood). These patients reported a wide variety of comorbidities, particularly anxiety disorders and dysthymia.

Of the total, 38% of patients with data in the system (2255/5998) reported severe lack of motivation and 28% (2889/10,372) reported severe difficulty in concentrating. Among patients with severe depression, those reporting severe difficulty in concentrating were approximately five-fold more in number than those reporting no/mild depressive symptoms (Figure 2). Notably, persistent difficulty in concentration was common in all severities of depression and reported by as many as 80% of patients, even in the absence of depressed mood, or with self-described “mild” depression (Figure 2).

The results of the retrospective database analysis also revealed that across 12 of the most frequently reported medications, only approximately 20% of patient reports (n=3095) indicated major effectiveness of the antidepressant treatment. Of the patients who considered themselves successfully treated, 24% (124/521) and 17% (149/859) still experienced a severe lack of motivation and difficulty in concentration, respectively.

The study 1 results highlighted three key topics to be included in the study 2 survey questionnaire: patient prioritization of symptoms, patient perception of cognitive symptoms of depression, and patient perception of residual symptoms and medication effectiveness.

Regarding patient prioritization of symptoms, patients reported lack of motivation and problems in concentrating as their most severe symptoms while experiencing a severely depressed mood. For some patients, these suggested mood and cognitive symptoms could be a higher priority for symptom resolution than physical symptoms (eg, fatigue and muscle tension). Therefore, the first key objective for study 2 was to understand how patients prioritize resolution of symptoms and how they define their own depression.

With regard to patient perception of cognitive symptoms of depression, study 1 results indicated that the majority of patients suffered some level of impairment in concentration, even when they reported no or mildly depressed mood. However, no data were available to assess other domains of cognition symptoms, such as executive function, attention, and memory; therefore, the second key objective of study 2 was to understand patient perceptions of cognitive symptoms related to depression, of these symptoms on their disorder, their daily life, and as symptoms independent of mood symptomatology.

Study 1 also suggested that many patients experienced residual symptoms when receiving a treatment that was perceived to be effective, including severe lack of motivation and problems concentrating. However, PatientsLikeMe data do not distinguish between residual symptoms of depression and treatment-related side effects. Thus, study 2 aimed to identify residual symptoms that were most important to patients, expectations of symptom resolution, and factors that influence perception of medication efficacy.

The questionnaire survey sample included 525 members with MDD. Survey respondents had demographic characteristics similar to those of the “total PatientsLikeMe MDD community” (Table 2). As in study 1, the majority of participants were women (74%), white (87%), and living in the United States (95%), and 90% had at least some college education, undergraduate degree, or graduate degree.

The characteristics of survey participants are shown in Table 2. PHQ-8 scores indicated moderate (9%), moderately severe (20%), and severe (67%) rates of depression, with a mean score of 23 (SD 6.0). A total of 44% of participants reported a current episode with continued symptoms. In addition, 71% of patients reported at least five past depressive episodes, and 87% reported currently taking medications for MDD. In total, according to data from the PatientsLikeMe platform database, 79% of the survey participants (416/525) reported more than three comorbid conditions—most commonly, generalized anxiety disorder, posttraumatic stress disorder, panic disorder, and dysthymia. The most-common nonpsychological conditions were fibromyalgia, hypertension, hypothyroidism, and gastroesophageal reflux disease.

When patients were asked whether they were able to tell if individual symptoms were improving, 73% of patients (383/525) agreed that they could distinguish between different symptoms of their depression and discern improvements in individual symptoms. In addition, although 72% of respondents (364/508) stated that the symptoms they most wanted to “go away or control” were mood and physical symptoms (depressed mood, anhedonia, anxiety, fatigue, insomnia, muscle tension, feelings of guilt or inadequacy, and low self-esteem), 23% (118/508) cited that the most important symptoms were cognitive symptoms (difficulty staying motivated, difficulty staying focused on goals, concentration problems, short-term memory problems, inability to make decisions, and attention problems).

Overall, 83% of respondents (433/523) reported speaking with their doctor about cognitive problems, and when asked to describe the experience (via free-text responses), 73% of the respondents (172/237) reported that they had received advice or treatment. The mean PDQ-5 score of respondents was 13.2 (SD 4.2), with 31% of respondents (165/525) scoring a higher level of perceived cognitive impairment (scores of 16-20).

Most patients attributed their cognitive difficulties to a combination of factors, including MDD (82%, 432/524) and other health (eg, fibromyalgia and Parkinson disease; 74%, 390/524) and life (eg, stress; 84%, 439/524) factors; only 6% (30/524) attributed these symptoms solely to MDD.

Many patients reported cognitive symptoms other than those recorded in the PDQ-5. The cognitive symptoms that were most strongly correlated with severity of depression, as measured by the PHQ-8, were difficulty making decisions, concentrating, and thinking clearly (Spearman rank correlation coefficients [r_s] of 0.32, 0.36, and 0.34, respectively; Figure 3).

When asked to select three symptoms with the most negative impact, depressed mood was the most frequently reported symptom to have a negative impact on family and social relationships (39%, 196/509) and fatigue had the most negative impact on social and leisure activities (43%, 219/510). Problems with memory, motivation, and concentration/attention also interfered with these aspects of life. In all, 66% of respondents stated that cognitive difficulties interfered with their ability to have meaningful relationships, and 69% reported that cognitive difficulties impaired their ability to maintain the household (Figure 4). More than half the patients reported that memory, thought, and concentration difficulties reduced their ability to handle household finances (54%), plan and serve meals (53%), and take care of themselves (54%; Figure 4a). When asked to provide free-text responses about how cognitive symptoms impacted them in any way that was not previously asked, common response themes among the respondents included interference with work/school (34%), disrupted relationships (28%), personal issues (27%), and communication problems (12%; Figure 4a).

Cognitive difficulties also had a profoundly negative impact on patients’ ability to work. In all, 35% of respondents cited poor concentration as one of the three symptoms with the most negative impact on occupational ability, compared with only 22% who cited symptoms of depressed mood (Multimedia Appendix 2). In total, 65% of respondents (341/525) reported that memory, thought, and concentration difficulties interfered with their ability to work effectively. A total of 48% of patients had stopped working completely as a result of the disability caused by these symptoms, and only a minority were employed full time (18%, 88/498) or part time (8%, 41/498).

In the free-text responses, patients commonly used the terms “brain fog,” “memory,” and “remember” when describing their cognitive symptoms. Free-text analysis revealed themes of feeling paralyzed, debilitated, and frustrated by symptoms:

Patients reported that cognitive symptoms “often” or “almost always” interfered with their ability to recover fully from depression (24% [125/525] and 19% [98/525] respondents, respectively), prevent relapse (22% [116/525] and 16% [84/525], respectively), and participate actively in treatment (15% [78/525] and 8% [43/525], respectively). Therefore, patients recognized cognitive difficulties as a distinct entity from other symptoms of depression.

Most patients reported multiple lifestyle adjustments to try to overcome the challenges posed by their cognitive symptoms. These included using reminders (eg, electronic organizer, checklists, and sticky notes; 85%, 439/516), organizing their living space (49%, 255/516), and taking public transportation instead of driving themselves (15%, 79/516). Patients stated that relief from emotional symptoms (91%; 454/501 respondents) and relief from cognitive symptoms (75%, 378/501) were “very important” in treatment decisions; in addition, a number of patients reported relief from emotional symptoms (7%, 37/501) and cognitive symptoms (19%, 96/501) as “somewhat important.” However, 61% of respondents (319/525) considered antidepressant treatment successful as long as their most troublesome symptoms improved. Overall, 9% of the respondents (46/501) stated that “the need to think more clearly” was the main priority in selecting antidepressant treatment, compared with 17% (85/501) who cited low rates of side effects and 13% (64/501) who were guided primarily by HCP recommendations.