A Patient-Centered Perspective on Changes in Personal Characteristics After Deep Brain Stimulation

Key Points Question Is deep brain stimulation (DBS) associated with unwanted changes in personally valued characteristics among patients with Parkinson disease (PD)? Findings This cohort study of 49 patients with PD and their care partners studied changes in visual analog scale ratings reflecting the personal characteristics that patients most feared losing. Scores of the patients and care partners on the visual analog scales increased over time, reflecting greater manifestations of (ie, positive changes in) valued characteristics following DBS. Meaning These findings may be relevant to informing decision-making for patients with advanced PD who are considering DBS.

The current study focused on the surgical cohort who was part of a larger study assessing how a large sample of patients with PD and their care partners (n=156 dyads) describe the patients' personality over the course of their lifetime to better capture their understandings of personality.We operationally defined personality in our interviews as characteristic ways of thinking, feeling, and behaving per one of the classic psychological definitions.A subset of transcripts (n = 60, including 20 transcripts from our surgical cohort reflecting the first 10 dyads) was reviewed, and broad themes were identified.Responses to the question related to the patient's personality over their lifetime underwent thematic content analyses to inductively and iteratively identify recurring themes related to participants' personality descriptions.Transcripts were read closely again to identify more nuanced and specific codes that fell within the larger themes. 32This process resulted in an initial set of 32 codes (e.g., "loving, caring", "happy, optimistic, positive", see eTable 1).
After this coding structure was finalized, two raters independently coded a subset of 60 transcripts (30 patient, 30 care partners from the total sample of 156 surgical participant-care partner dyad transcripts) balanced by interviewer, age, gender, and race/ethnicity.543 unique descriptors (e.g., "loving", "positive", "happy") were double coded from this data subset.Kappa statistics were computed to determine the inter-rater reliability of the coding structure.Excellent inter-rater reliability was observed (Kappa=.80).
To simplify data presentation and identify the underlying factor structure of personality as defined by our sample, we utilized an exploratory factor analysis (EFA).The care partner participants' descriptions of the patients' personality (n=156) were coded using the previously identified codes by one coder (CSK).Each excerpt of text (total = 886) from the care partner interviews describing the patient's personality was transformed into a survey item with a four-point Likert scale with "Not at all similar" at one end and "Extremely similar" at the other.Two raters independently rated the similarity of every text excerpt to each of the 32 codes derived from the content analysis using the four-point Likert scale.Interclass correlations were conducted to assess reliability of ratings.Averages for the two raters were then computed for each personality code that was determined to be reliable between raters.
Those averages were used as variables in an exploratory factor analysis (EFA) to examine the relationships among the personality codes and identify higher level codes.
The EFA indicated that an eight-factor solution fit the data best.χ 2 (268) = 1776.308,p = .000,RMSEA = 0.08, CFI = 0.865, TLI= 0.750, SRMR = 0.032.Table S1 provides the individual 32 codes derived from the content analysis and the factor with which they loaded highest.As noted in the main text, we attempted to label the factors with terms commonly used in the psychological literature (e.g., internalizing vs externalizing negative emotionality like internalizing vs externalizing disorders).We acknowledge that traditional personality tests do not incorporate physical abilities or characteristics; however, these were characteristics our sample identified in their descriptions of personality and correspond to other considerations of self in which physical (i.e., motor, sensory) characteristics are incorporated. 22We also recognize that some of the personality codes that loaded on the Mixed Character "Virtues" factor may not correspond to classically defined virtues and that other factors (e.g., Prosocial, Self-Regulation) may contain personality codes traditionally included in classic discussions of virtue.For the purposes of this initial paper, we opted to use the descriptive term Mixed Character "Virtues" as many of the codes in this factor correspond to what the public would characterize as virtuous behavior.

eMethods 1 .
Interview Protocol eMethods 2. Qualitative Methods and Results eTable 1. Factors and Related Personal Characteristic Codes eTable 2. Changes in Mean Top 3 and Global VAS Ratings Over Time eFigure 1. Mean Patient VAS Scores for Top 3 Characteristics Over Time With 95% CI eFigure 2. Mean Care Partner VAS Scores for Top 3 Characteristics Over Time With 95% CI eFigure 3. Association Between Patient and Care Partner Participants' Scores Over Time Across All 3 Characteristics eFigure 4. Mean Top 3 Characteristic Scores With 95% CI This supplementary material has been provided by the authors to give readers additional information about their work.eMethods 1. Interview Protocol We are interested in better understanding any changes you may have observed over the course of your Parkinson's disease in your personality.Psychologists define personality as characteristic patterns of thinking, feeling, and behaving.We are particularly interested in your perceptions regarding changes in who you are as a person.Please take a few minutes and tell me some of the words that you feel best describe your personality throughout most of your life.…….. Great.Now, I'd like you to tell me the top five personality traits that you most fear losing.(*Some participants could not generate five; thus, we required a minimum of three) Please rank order these characteristics such that #1 is the feature you most fear losing (most important to maintain)mark on the line below indicating the extent to which you demonstrate Trait 1 ________________ at the present time.Zero represents the least manifestation of that trait whereas 10 represents the most extreme manifestation of that trait.mark on the line below indicating how much you feel you are the same person you have always been with respect to your thoughts, feelings, and actions since you noticed your first symptom of Parkinson's disease.
Association Between Patient and Care Partner Participants' Scores Over Time Across All 3 Characteristics 6. eFigure 4. Mean Top 3 Characteristic Scores With 95% CI