Assessing therapeutic change in patients with severe dissociative disorders: the progress in treatment questionnaire, therapist and patient measures

ABSTRACT Background: Treatment research for dissociative identity disorder (DID) and closely related severe dissociative disorders (DD) is rare, and has been made more difficult by the lack of a reliable, valid measure for assessing treatment progress in these populations. Objective: This paper presents psychometric data for therapist and patient report measures developed to evaluate therapeutic progress and outcomes for individuals with DID and other DD: the Progress in Treatment Questionnaire – Therapist (PITQ-t; a therapist report measure) and the Progress in Treatment Questionnaire – Patient (PITQ-p; a patient self-report measure). Method: We examined the data of 177 patient–therapist pairs (total N = 354) participating in the TOP DD Network Study, an online psychoeducation programme aimed at helping patients with DD establish safety, regulate emotions, and manage dissociative and posttraumatic symptoms. Results: The PITQ-t and PITQ-p demonstrated good internal consistency and evidence of moderate convergent validity in relation to established measures of emotional dysregulation, dissociation, posttraumatic stress disorder, and psychological quality of life, which are characteristic difficulties for DD patients. The measures also demonstrated significant relationships in the hypothesized directions with positive emotions, social relations, and self-harm and dangerous behaviours. The patient-completed PITQ-p, which may be used as an ongoing assessment measure to guide treatment planning, demonstrated evidence of stronger relationships with established symptom measures than the PITQ-t. Conclusions: The PITQ-t and PITQ-p merit use, additional research, and refinement in relation to the assessment of therapeutic progress with patients with DD.

and 1% of all available community samples internationally, and that prevalence increases with level of care (median outpatient prevalence rate: 2.5%, median inpatient prevalence rate: 5%; Foote, 2016). Dissociative disorder not otherwise specified (DDNOS), the DSM-IV-TR diagnosis assigned to patients with symptoms similar to DID (Spiegel et al., 2011), is more than twice as prevalent in the U.S. (4.4%; Johnson et al., 2006). These severe dissociative disorder (DD) patients frequently present with myriad difficulties, including emotion dysregulation, intrusions of traumatic experiences and other symptoms of posttraumatic stress disorder (PTSD), dissociation, self-harm, suicidality, attachment and relationship difficulties (including mistrust of therapists), and poor quality of life (Brand, Classen, Lanius et al., 2009;Foote, Smolin, Kaplan, Legatt, & Lipschitz, 2006;Johnson et al., 2006;Liotti, 1992;Mueller-Pfeiffer et al., 2012;Pasquini, Liotti, Mazzotti, Fassone, & Picardi, 2002). Treatment research for DD patients is scarce, however (Brand, Classen, McNary, & Zaveri, 2009), and is made more difficult by the lack of a reliable, validated measure for assessing DD treatment progress. To address this need, this paper presents psychometric data for two brief measures for assessing DD treatment progress and outcomes: the Progress in Treatment Questionnaire -Therapist (PITQ-t; a therapist report instrument), and the Progress in Treatment Questionnaire -Patient (PITQ-p; a patient self-report instrument). These measures represent new versions of the Progress in Treatment Questionnaire (PITQ; Brand, Classen, Lanius et al., 2009), a therapist report measure designed to assess DD patient capacities targeted in the treatment of adult DID patients.

The Progress in Treatment Questionnaire
The Progress in Treatment Questionnaire (PITQ) was developed for the naturalistic Treatment of Patients with Dissociative Disorders (TOP DD) study (Brand, Classen, Lanius et al., 2009. The TOP DD study was a 30-month naturalistic study of DD patients' progress in individual therapy. Prior to this study, only one DID-specific progress/outcome measure had been developed: the Dimensions of Therapeutic Movement Instrument (DTMI; Kluft, 1994). The DTMI is completed by clinicians who rate their DD patients in 12 areas, including the therapeutic relationship, self-regulation, need for external support, phenomena related to dissociative self-states (DSS), 1 interpersonal functioning, and sense of well-being. The DTMI offers promise as a therapist-completed assessment measure, but has not been systematically assessed for reliability or validity (Choe & Kluft, 1995;Kluft, 1994) and is sufficiently complex that it was deemed too difficult to use in online research such as the TOP DD study.
For these reasons, the TOP DD investigators developed the PITQ. As described in Brand, Classen, Lanius et al. (2009), the DD treatment experts involved in the study designed the therapist-completed PITQ by consensus to assess DD patients' ability to safely and effectively manage their emotions, symptoms, and relationships; these are adaptive capacities developed across the three stages of treatment described in the International Society for the Study of Trauma & Dissociation's (ISSTD) Guidelines for Treating Dissociative Identity Disorder in Adults (2005Adults ( , 2011. PITQ item development was informed by the ISSTD Treatment Guidelines, the items in the DTMI, and the TOP DD team's extensive clinical experience. PITQ items most closely linked with the first stage of treatment, establishing patient safety and stabilization ISSTD, 2011), query patients' ability to maintain safety (i.e. not engage in behaviours that put their life or health at risk, including non-suicidal self-injury, suicidality, and other dangerous behaviours); effectively collaborate with their therapist; understand and manage DD-and PTSD-related symptoms; manage self-care/ daily living; view themselves and others in realistic ways that are free from trauma-based distortions; maintain healthy personal and professional relationships; tolerate, regulate, and identify affect; and develop cooperation with DSS. PITQ items most closely linked with the second phase of treatment, processing trauma, query patients' ability to process trauma and identify and understand the roles of DSS. PITQ items most closely linked with the third phase of treatment, reconnection, query patients' ability to resolve (Kluft, 1993;Kluft & Loewenstein, 2007) the conflicts between and/or integrate DSS and find ways to make life feel meaningful and rewarding. An item also queries patients' ability to pleasurably experience sexual intimacy, which can be difficult for individuals with DDs. The PITQ's 29 items ask therapists to indicate what percentage of the time their patient has demonstrated the queried abilities over the prior six months. Each item offers 11 response options ranging from 0% (never) to 100% (always) in 10% intervals (i.e. 0%, 10%, 20%, etc.).
1.1.1. Psychometric properties of the PITQ 1.1.1.1. Reliability. The PITQ has demonstrated evidence of internal consistency (i.e. internal reliability) (Brand, Classen, Lanius et al., 2009Marsic, Brand, Schielke, & Putnam, 2013) and test-retest reliability (Marsic et al., 2013) in studies with DD patients. In a 30-month prospective, naturalistic study of the treatment of DD patients, the PITQ demonstrated Cronbach's coefficients ranging from .94 to .95 (see Brand, Classen, Lanius et al., 2009, andBrand et al., 2013 for methodology details). In an IRB-approved study aimed at assessing the test-retest reliability of the PITQ, we invited therapists who had participated in the naturalistic study to complete the PITQ at two time points (two weeks apart) for a single adult outpatient diagnosed DID or DDNOS by DSM-IV-TR criteria (Marsic et al., 2013). To increase generalizability, we did not exclude patients based on psychiatric comorbidity. No patient-identifying information was collected, and therapist participants' data was de-identified. Participation was excluded for therapists without access to the internet or unable to read English. A total of 117 therapists consented to participate in the study; 74 (53 females and 21 males) completed the PITQ at both Time 1 and Time 2 (retention rate: 63%). Responses were matched by participant code numbers. Analysis of participants' test responses revealed a strong Pearson's product moment correlation (.87, p > .01) between the two administrations, suggesting good test-retest reliability and temporal stability over a two-week period (M = 16.6 days). The probability of these findings occurring in the same population by chance was found to be <1% [t(73) = −2.72, p = .008], and a Pittman-Morgan test of homogeneity of variance indicated that there was no significant difference in variance between the two samples [t(72) = .61, p = .54] (SD Time 1 = 16.09, SD Time 2 = 14.99). The PITQ demonstrated good internal consistency at both time points (Cronbach's alpha: .95 and .94, respectively).
1.1.1.2. Validity. Abilities assessed by the PITQ correspond with DD treatment experts' endorsed intervention targets (see Table 1). As a measure intended to assess patients' development of adaptive capacities in managing safety, emotion, symptoms, and relationships, PITQ scores should be higher in patients in later stages of treatment. Consistent with this theory, in cross-sectional comparisons, mean PITQ scores were significantly higher in patients in the 'processing trauma' and 'reconnection' stages than in patients in stage 1 (p < .01; Hedge's g, 'processing' vs. 'establishing safety': 1.73; Hedge's g, 'reconnection' vs. 'establishing safety': 2.90; . Similarly, in longitudinal analyses of TOP DD participants over 30 months, PITQ scores increased over time ; these increases in PITQ scores coincided with reductions in symptoms of dissociation, depression, general distress, and PTSD, as well as with reductions in substance use and dangerous behaviour.

PITQ-t development
The therapist-completed PITQ-t is a modestly updated version of the original PITQ. To explore use of the PITQ-t with patients that do not report DSS, we reorganized the measure such that DSS-related items are presented last, preceded by instructions to only rate DSS functioning if a patient experiences DSS. In response to suggestions that we clarify three PITQ items related to patient difficulties, we re-phrased and reverse-scored these items (Table 5, items 1-3). The remaining 26 items, response options, and targeted time frame are unchanged from the original PITQ.

PITQ-p development
Recent research underscores the value of querying patients' views about their treatment. A meta-analysis of six studies comprising over 6000 treatments (Shimokawa, Lambert, & Smart, 2010) found that regularly assessing and reviewing patient treatment progress feedback with the OQ patient feedback system (e.g. Lambert, 2015) resulted in improved outcomes for patients demonstrating difficulties in making or sustaining treatment gains. A meta-analysis of 201 therapeutic alliance studies comprising over 14,000 treatments (Horvath, Del Re, Flückiger, & Symonds, 2011) found that patient-reported alliance data were generally more strongly predictive of therapeutic outcome than therapist-reported data. These findings prompted the first and second authors to develop a patient version of the PITQ. We believed that a patient version of the PITQ could afford additional insight into patients' therapeutic progress and inform treatment planning while also offering patients opportunities to reflect on how they are doing in relation to DD treatment targets. Informed by the therapeutic alliance meta-analysis, we also anticipated that such a measure would demonstrate a stronger relationship with variables indicative of therapeutic progress than the therapist-completed PITQ-t.
In the 32-item PITQ-p, DSS items are presented last, preceded by instructions to complete these items only if relevant. Two PITQ prompts related to fusion and integration of DSS were not included because of the potential for some patients to perceive these as suggesting DSS 'loss' or 'death' (Kluft, 1993). We rephrased the remaining 27 of the PITQ's 29 items in patient-appropriate language, and added five items (items 3, 4, 12, 28, and 31 in Table 7) in the interest of raising patient awareness of (and tracking progress related to) these adaptive capacities. To increase the PITQ-p's potential usefulness as a routinely-administered measure, we also changed the targeted time frame to 'in the last week'.

Method
We examined the internal consistency and convergent validity of the PITQ-t and PITQ-p in use with DD patients at baseline in a prospective longitudinal study. We hypothesized that higher PITQ-t and PITQ-p scores would be negatively correlated with concurrent validated measures indicating difficulties with emotional dysregulation, PTSD, and dissociation symptoms. We also hypothesized that PITQ-t and PITQ-p scores would be negatively correlated with self-harm and dangerous and impulsive actions, and positively correlated with positive emotions and quality of life.

Participants
Participants consisted of 177 patient-therapist pairs (total N = 354 individuals) who consented to participate in the TOP DD Network study, an IRB-approved study examining the effectiveness of an online psychoeducational programme aimed at helping early-stage DD patients learn skills to help them regulate emotion, manage DD and PTSD symptoms, and establish safety from behaviours that put their life or health at risk, including non-suicidal self-injury, suicidality, and other dangerous behaviours. We solicited participants through flyers and announcements at trauma and dissociation-focused conferences, email invitations to trauma-and dissociationrelated mailing lists, and emails to therapists who had taken part in the naturalistic TOP DD study. Therapists were asked to invite a patient to participate who met diagnostic criteria for Dissociative Identity Disorder (DID; DSM-IV-TR, DSM-5), Dissociative Disorder Not Otherwise Specified (DDNOS; DSM-IV-TR), or Other Specified Dissociative Disorder (OSDD; DSM-5); had been in treatment with them for at least three months; was 18 years or older; had reliable high-speed internet access; was able to read English at the eighth grade level; could tolerate non-detailed references to trauma, safety struggles, dissociation, and brief mentions of 'parts of Table 1. Relationships between abilities assessed by PITQ-t, PITQ-p, and expert intervention targets.
Interventions that experts endorse using 'very often' during at least one stage of treatment a PITQ-t items assessing ability targeted by intervention b  Brand et al. (2012). b PITQ-t items 4-29 (see Table 5 for item content) are taken verbatim from PITQ items; PITQ-t items 1-3 rephrase the content of the remaining 3 PITQ items for clarity. c See Table 7 for PITQ-p item content. * Not a patient-targeted ability.
PITQ family of measures does not explicitly query, but is very closely related to (items in parentheses).
self', even if this term did not apply to them; and was willing to do approximately 2½ hours of study-related activities per week. No patients were excluded on the basis of co-morbid disorders, self-harm, or suicidality.
For the purposes of the present study, which analyses baseline data gathered prior to participants' engagement with the Network psychoeducational programme, we only included participant pairs that unanimously indicated that the patient had DSS. See Tables 2 and 3 for therapist and patient characteristics, respectively.

Progress in Treatment Questionnaire, therapist version (PITQ-t)
The PITQ-t (available at TOPDDStudy.com/PITQ-t) is a therapist-completed 29-item measure of dissociative patients' ability to manage their emotions, symptoms, relationships, safety, and well-being. The instrument prompts therapists to assess the percentage of time their patients demonstrate expert-identified DD treatment-related behaviours over the prior six months, and includes six items completed only for patients with DSS (items 24-29). Response options range from 0% (never) to 100% (always); the first three items are reverse-scored. Higher scores indicate better adaptive functioning.

Progress in Treatment Questionnaire, patient version (PITQ-p)
The PITQ-p (available at TOPDDStudy.com/PITQ-p) is a patient self-report measure that assesses DD patients' ability to manage their emotions, symptoms, relationships, safety, and well-being. The PITQ-p consists of 32 expert-identified items; items 27-32 are only completed by patients who report experiencing DSS. Patients are instructed to report what percentage of time they demonstrated a behaviour during the prior week, ranging from 0% (never) to 100% (always).
Higher scores indicate better adaptive functioning.

Difficulties in emotion regulation scale (DERS; Gratz & Roemer, 2004)
The DERS is a 36-item self-report measure that assesses six domains associated with emotion regulation: emotional acceptance; emotional awareness; ability to identify and understand emotions; access to effective selfregulation strategies; ability to control impulses; and ability to engage in goal directed behaviour when upset. Response options range from 1 (almost never; 0-10%) to 5 (almost always; 91-100%); higher scores indicate greater difficulty with emotion dysregulation and poorer psychological functioning. The DERS has high internal consistency (α = .93), good overall testretest reliability (ρI = .88, p < .01) and adequate subscale test-retest reliability (ρIs ranging from .68 to .89, all ps < .01) over 4-8-weeks, and adequate construct and  predictive validity (Gratz & Roemer, 2004 The PCL-C is a 17-item self-report measure that assesses DSM-IV-TR PTSD symptoms (APA, 2000).
Patients rate the frequency of experiencing each symptom over the past month from 1 (not at all) to 5 (extremely). Higher scores indicate greater PTSD symptoms (Weathers & Ford, 1996). The measure has high overall diagnostic efficiency (0.90; Blanchard, Jones-Alexander, Buckely, & Forneris, 1996) and strong test-rest reliability (.96 with 2-3 day interval; Weathers et al., 1994). In this sample, Cronbach's alpha was .88.  Table 4 for items' wording and descriptive statistics.

Procedure
The study was approved by the Institutional Review Board at Towson University. Survey data were collected using secured, password-protected websites. All participants completed informed consent forms prior to participating. Participants were not compensated for participation.

Analysis strategy
All analyses were performed on baseline data (i.e. data collected prior to participants engaging with the Network psychoeducational programme). To evaluate PITQ-t and PITQ-p internal consistency, we calculated Cronbach's alphas and examined itemtotal statistics to assess whether internal reliability would be meaningfully improved by item deletion.
To examine convergent validity, we assessed correlations between the PITQ-t and PITQ-p and the DERS, DES II, and PCL-C, as well as with patient reported self-harm, dangerous and impulsive actions, positive emotions, and psychological and social quality of life. Since higher scores for the DERS, DES II, and PCL-C are associated with greater psychological difficulty (in contrast to the PITQ-t and PITQ-p), we hypothesized that we would find negative correlations between concurrent administrations of these measures and the PITQ-t and PITQ-p. We also hypothesized that PITQ-t and PITQ-p scores would be negatively correlated with patient self-harm and dangerous and impulsive actions, and positively related to positive emotions and psychological and social quality of life. Informed

PITQ-t
3.1.1. Internal consistency Cronbach's alpha (.91) indicated that the PITQ-t demonstrated good internal consistency. Corrected item-total correlations (see Table 5) generally ranged from .33 to .69, with three exceptions: Item 2, 'Engages in potentially self-damaging acts such as abusing substances, purging, shoplifting, driving unsafely' (reverse scored) demonstrated a corrected item-total correlation of .24; Item 3, 'Identity is strongly tied to being a victim of abuse' (reverse scored) demonstrated a corrected item-total correlation of .09; and Item 22, 'Able to experience sexual intimacy without difficulties such as intense shame, flashbacks or dissociation and with some pleasure' demonstrated a corrected item-total correlation of .23. Examination of item-total statistics indicated that Cronbach's alpha would be unchanged by deleting items 2 or 22, and would become .92 if Item 3 were deleted. Given the clinical salience of these items for DD patients, and that their deletion would have little impact on internal consistency, we retained them.

Convergent validity
Correlations between the PITQ-t and the DERS, DES II, and PCL-C were significant at the .01 level (2-tailed) in the hypothesized negative direction: higher scores on the PITQ-t were associated with lower scores on the DERS (r = -.35), DES II (r = -.29), and PCL-C (r = -.41) (see Table 6). Correlations between the PITQ-t and unsafe behaviour in the last 30 days were also significant (p < .01) in the expected negative direction: self-harm (r = -.37), dangerous behaviour (r = -.31), and impulsive behaviours (r = -.23). Finally, the PITQ-t demonstrated the expected positive correlations (p < .01) with psychological (r = .45) and social (r = .22) quality of life, and positive emotions (r = .21) (see Table 6).

Internal consistency
The PITQ-p also demonstrated good internal consistency (α = .92). Corrected item-total correlations (see Table 7) generally ranged from .36 to .64, with two exceptions: Item 1 ('I have been diagnosed with a dissociative disorder and agree that this diagnosis is correct') demonstrated a corrected item-total correlation of .17 and Item 25 ('I am able to experience sexual intimacy without intense shame, flashbacks, or dissociation, and with some pleasure') demonstrated a corrected item-total correlation of .27. Examination of item-total statistics for these items indicated that reliability would not be improved by deleting either item.

Comparative strength of relationships between PITQ-t and PITQ-p and outcomes
Results of Williams' standard t (see Table 6) indicated that the differences between the PITQ-t and PITQ-p's correlations with outcomes were significant only in relation to the DERS (t 174 = 5.51, p < .001) and WHOQOL-BREF Psychological Domain (t 174 = −3.21, p < .001). The PITQ-p demonstrated evidence of significantly stronger correlations than the PITQ-t with each of these measures.

Discussion
Treatment outcome research for DD patients is scarce. The lack of a reliable, validated measure for assessing progress towards resolving the myriad difficulties faced by DD patients contributes to the paucity of treatment research. To address this gap, we developed therapist (PITQ-t) and patient (PITQ-p) measures of the adaptive capacities expected to develop in DD patients over the course of trauma-and dissociationfocused treatment and examined their internal consistency and construct validity. The PITQ-t and PITQ-p demonstrated evidence of good internal consistency, and significant, generally moderate concurrent relationships with established measures of adaptive emotion-related functioning, PTSD symptoms, dissociation, and psychological and social quality of life. Higher PITQ-t and PITQ-p scores were also related to lower levels of self-harm, dangerous behaviours, and impulsivity, as well as higher positive emotions. This suggests that higher scores on the PITQ-t and PITQ-p are associated with better functioning in each of these domains, lending support to the use of the PITQ-t and PITQ-p in evaluating DD patients' ability to manage their emotions, symptoms, relationships, safety, and well-being.
Consistent with our rationale for developing the PITQ-p, we found that the patient-completed measure demonstrated stronger relationships with established symptom measures than the therapistcompleted PITQ-t. Specifically, the correlations between the PITQ-p and measures of dissociation, PTSD, and emotion dysregulation were greater than those between these measures and the PITQ-t. The PITQ-p also demonstrated statistically stronger relationships with measures of emotional dysregulation and psychological quality of life than the PITQ-t. These findings are consistent with research indicating a stronger relationship between patients' reports (vs. therapist reports) of alliance and treatment outcome (Horvath et al., 2011), and underscore the salience of querying patients' perceptions about their progress in treatment.  Table 4. * p < .05 (2-tailed), ** p < .01 (2-tailed), r tp = correlation between PITQ-t and PITQ-p.
Chronic, complex DD have been conceptualized as disorders of affect regulation related to trauma and attachment difficulties (Brand & Lanius, 2014), and difficulties in affect regulation and posttraumatic stress have been found to predict increased dissociation (Briere, Hodges, & Godbout, 2010). Our findings add to the literature indicating the importance of assessing these areas of functioning in DD patients. The PITQ-p and PITQ-t assess adaptive capacities related to these difficulties, and the TOP DD naturalistic study demonstrated that these capacities develop over time in DD treatment (Brand, Classen, Lanius et al., 2009.

Limitations and areas for future study
This self-selected sample population consisted predominantly of female Caucasian patients that presented with DSS seeing primarily female therapists.
Future studies should strive to increase representation of groups under-represented here, examine the validity and utility of the measures with patients who do not demonstrate or endorse DSS, and further examine the PITQ-t and PITQ-p's consistency and validity in terms of replication and extension. This includes examining the measures' temporal stability, discriminant validity, and utility with DD patients as well as with other populations that experience dissociation, including patients with the dissociative subtype of PTSD (DSM-5;Lanius, Brand, Vermetten, Frewen, & Spiegel, 2012;Stein et al., 2013;Wolf et al., 2012) and/or those reporting complex trauma histories (Cook et al., 2005;Courtois, 2004;Ford & Courtois, 2009). Given that the measures studied in relation to the patient self-report version of the PITQ were each selfreport measures, it is possible that shared method Items 3, 4, 12, 28, and 31 reflect capacity targets introduced in the PITQ-p.
variance may be meaningfully contributing to the PITQ-p's stronger relationships with these measures. Future studies could include additional observational and/or collateral outcome measures as well as measures of divergent constructs and examine this question through the multi-trait, multi-method matrix approach (Campbell & Fiske, 1959). The measures' relationships with other instruments and variables of interest should also be examined over multiple time-points within a longitudinal study, a process that is currently underway. Future studies could also examine if completing the PITQ-p during the course of therapy helps patients acquire awareness of the links between adaptive abilities and improved psychological quality of life more rapidly than those that do not complete the measure, and could examine whether regular use of the PITQ-p improves therapeutic progress and/or outcomes. Finally, in the present study, patients completed single-item variables that assessed self-harm and impulsive and dangerous behaviours over the prior 30 days. Future studies should match the time frame for patient-reported outcomes with the seven-day time frame used by the PITQ-p.

Conclusions
Our findings suggest that the PITQ-t and PITQ-p are promising measures of adaptive functioning in DD patients and merit use and additional research in relation to the assessment of therapeutic progress with patients with DD. These measures seem to capture important aspects of DD patients' development of adaptive capacities in managing safety, emotion, symptoms, and relationships, and offer researchers and clinicians a means of tracking progress and studying the process of therapeutic change with this understudied population. They are also promising tools for evaluating patients' relative strengths and areas of difficulty, information that can guide treatment planning (Lambert, Gregersen, & Burlingame, 2004;Pinsof & Chambers, 2009) and facilitate treatment responsiveness (Stiles, Honos-Webb, & Surko, 1998).
Therapists have informed us that these measures have been useful in treatment planning and in discussing treatment progress with patients. The PITQ-p may be useful in facilitating treatment responsiveness by helping clinicians efficiently evaluate multiple areas of their DD patients' functioning. Asking patients to periodically complete a PITQ-p enables therapists to rapidly assess the relative presence or absence of a range of expert-identified abilities to be targeted in treatment, allowing therapists to quickly assess areas of functioning that are frequently challenging for DD patients and providing important information that can facilitate treatment planning.
Completing the PITQ-p also encourages patients to reflect on their progress towards treatment goals, which may, in turn, encourage increased use of the skills assessed within the measure. Note 1. Dissociative self-states are sometimes referred to as dissociative identities, personalities, alters, or 'parts'.

Disclosure statement
No potential conflict of interest was reported by the authors.