A graduate student, two years under the wrong diagnosis, and the limits of an accurate answer.
The freezer
On the Saturday of a holiday weekend, a minus-eighty freezer in a molecular biology building lost power sometime after midnight, and by the time the alarm reached anyone who could act on it, fourteen months of Caroline’s samples had thawed. She was twenty-four, in the third year of a doctoral program, and she had labeled every one of those tubes herself in a system of her own design that nobody else in the lab was permitted to touch. She arrived within the hour and stayed for thirty-one more, re-racking tubes that could no longer be saved. She had not eaten since the previous evening and had run nine miles that morning, because it was Saturday and Saturday was nine miles.
A labmate drove her to the emergency department when her hands went numb and her chest tightened enough that she could not finish a sentence. The workup was the standard one and it came back clean, apart from a heart rate that settled into the forties once she calmed, which was read as athletic and was athletic as far as that explanation went. The attending told her, kindly, that she had experienced a panic attack and that she should follow up with psychiatry.
She corrected him. She explained that this was not anxiety, that she did not get anxious, and that what he had witnessed was the reasonable response of a person whose work had been destroyed by a facilities failure. She asked to see her troponin value. She was discharged with a referral and a prescription for a benzodiazepine that she never filled, on the grounds that she does not take things.
She was partly right, and it would take two more years and three diagnoses before anyone could tell her which part.
Two years of the wrong word
Student health saw her first and wrote generalized anxiety disorder after a twelve-minute appointment in which she explained twice that she was not anxious. The psychiatrist she reached next was a careful one, working inside the standard arrangement of a longer first visit followed by medication checks of fifteen or twenty minutes. Caroline described herself accurately and without distress. She arranged her bench and her kitchen in a fixed order, checked her labels three times, became upset when other people moved her things, and kept rules for most of the day.
Those sentences sound like obsessive-compulsive disorder to almost anyone, and order is the presentation of OCD that clinicians recognize most readily. In a vignette study of primary care physicians, half of all OCD presentations were misidentified, but the symmetry vignette was missed less than 4 percent of the time.1 That study tested the recognition of genuine OCD, so I cite it only as context for how quickly arranging and checking are heard as OCD.
She declined a serotonin reuptake inhibitor at the second visit and accepted a referral for exposure and response prevention, which is the correct referral for OCD and one of the best-evidenced treatments in the field. In the trial that anchors that literature, 62 percent of patients assigned to intensive exposure responded, against 8 percent on placebo.2
The therapist applied a standard OCD protocol in good faith, under the formulation she had been handed. She built a hierarchy on the assumption that disorder distressed Caroline, and she assigned her to leave a desk drawer unsorted for a week and to send an email containing a typographical error. Caroline completed every exercise, logged each one in a table she formatted herself, and reported no distress at any step. She treated the hierarchy as coursework, a pattern the exposure literature describes as the patient who tries to do the treatment perfectly and fixes on its technique while missing its purpose.3 Her self-report score on the Yale-Brown scale was 18 at intake and 16 five months later, and the chart began to carry the phrase treatment-resistant. The hierarchy was rebuilt twice and the exposures made harder, and for another year she completed those as well, declining the same prescription at each medication visit.
It was the psychiatrist, to her considerable credit, who said that the picture did not add up and asked for a full evaluation. Caroline agreed on the understanding that it would settle the matter, and she prepared for it the way she prepares for everything.
What the P changes
Obsessive-compulsive personality disorder was the most prevalent of the personality disorders assessed in the largest national survey of American adults, at 7.9 percent,4 and a quarter to a third of OCD patients also meet criteria for it,3 which is part of why the two blur.
The first study to compare them as separate clinical groups found both impaired and both showing compulsive behavior. Only the OCD patients reported obsessions, and only the OCPD patients showed an excessive capacity to delay reward, declining immediate money for larger sums later to a degree that exceeded healthy controls.5 It is an unusual disorder in which the laboratory measure of self-control is abnormal in the direction everyone admires.
The diagnostic line is drawn by what the repetitive behavior is for. OCD requires obsessions, compulsions, or both, and a compulsion is a behavior or mental act the person feels driven to perform, in response to an obsession or under rigid rules, to reduce distress or prevent a dreaded event, and that is clearly excessive or not realistically connected to what it is meant to prevent. The distress need not be fear. Many patients with ordering and symmetry symptoms repeat an act until an intolerable sense of incompleteness, of things being not just right, finally lets go.6 The structured interview asked Caroline about exactly that and did not find it. She arranges to a specification, stops when the specification is met, and does not repeat an act to settle a feeling. What she feels when someone moves her things is irritation at a broken standard, and she corrects it once. The manual describes OCPD as this kind of enduring and pervasive pattern of perfectionism and rigid control, as distinct from intrusive thoughts and the repetitive behaviors performed in answer to them.7
Clinicians often summarize the difference as ego-dystonic versus ego-syntonic, meaning that OCD is usually unwanted and OCPD usually feels like the self. The summary is useful and incomplete, since some patients with OCD have poor insight and will defend their rituals. Caroline does not regard her standards as symptoms, because from the inside they feel like the last functioning quality-control department in a careless world.
The treatment consequence is practical. Exposure and response prevention involves facing triggers, distress, or incompleteness without performing the compulsion, and Caroline’s routines were never compulsions in that sense. Exposure methods and behavioral experiments can be aimed at perfectionism and rigidity themselves, which is a different treatment with different targets.
A life that passes inspection
At twenty-six Caroline is a fifth-year doctoral candidate with a publication record her cohort envies and a second thesis chapter she has rewritten eleven times. Her advisor describes her as the most reliable person he has trained and has stopped asking when the chapter will be finished. First-year students are told on their first day not to touch her bench. She knows this, and in the evaluation she admitted, after some time, that she likes it.
Patients with OCPD and no OCD report hostile-dominant interpersonal problems and less empathic perspective taking than healthy controls,8 and Caroline would be surprised to learn that her labmates rehearse before asking her a question.
She dates the way she reviews manuscripts. Her longest relationship lasted eleven months and ended over the way he loaded a dishwasher, which she described to the evaluator as a difference in values and did not mean as a joke. There have been four first dates in the two years since and no second ones. The perfectionism literature calls this other-oriented perfectionism, the demand that other people meet one’s own exacting standard.9 No one is good enough for her for the same reason that she is not good enough for herself, which is that there is a single standard, applied without exemption, and set at a height where failing it is the ordinary condition. A man who loads a dishwasher badly is failing an inspection that she herself fails every day.
The one creature permitted on her bed is Franklin, a gray shorthair she adopted in college and named for a crystallographer. She weighs his food to the gram and speaks to him in a voice her sisters have not heard since she was small. As far as anyone can tell, he is the only member of any household she has lived in whom she has never tried to improve.
She has two sisters. Meredith is thirty-one, with two children and a house Caroline describes as loud. Paige is twenty-three, has changed majors twice, and tends bar. Caroline sends them articles on sleep and on sugar, and at Thanksgiving she reorganized Meredith’s pantry without being asked. She wants them closer, and the only instrument she has for closeness is correction. They keep a group text that does not include her, which she discovered by accident and has mentioned to no one. When Meredith was interviewed as a collateral source she cried and said she had always assumed that Caroline simply did not like them.
The arrangement runs on fear, which is the last thing anyone who knows her would guess. The people around her experience her as formidable. In the third session of the evaluation she described herself, in her own words, as always one mistake from being found out.
The cleanest diet in the building
Caroline stands five feet six and has weighed 116 pounds since she was nineteen, which is 1.4 pounds above the line where the word underweight begins, for a body mass index of 18.7. She weighs herself at 5:50 each morning, eats the same measured lunch every day, and consumes neither sugar nor anything she has not prepared herself. She objects to the word diet and prefers to say that she does not eat garbage.
She has run on 364 days of each of the last four years, the exception being the day after her annual half marathon, and two winters ago she ran for five weeks on a tibial stress reaction because the plan said to run. She has never taken a prescription medication and is not certain she has ever taken ibuprofen. She manages with magnesium and cold water, and she is prouder of this than of her publications.
The link between this temperament and disordered eating is well documented. In a case-control study, each additional obsessive-compulsive personality trait recalled from childhood multiplied the odds of a later eating disorder by nearly seven,10 and a systematic review of patients with anorexia nervosa found that excessive exercise tracked with these traits in most of the available studies.11
The evaluation did not diagnose an eating disorder, and it did not rule one out. A weight just inside the normal range and a denial of any fear of weight gain do not settle the question. Rigid restriction, a heart rate in the forties, daily running, and five weeks of training on a stress injury call for assessment in their own right, by a physician and by an eating disorder specialist, whatever the personality diagnosis turns out to be.
What the personality does explain is the moral weight she gives to all of it. One of the criteria for OCPD is inflexibility about matters of morality and values,7 and in Caroline that criterion has migrated into the body. Clean eating is a virtue and medication a weakness, which leaves people who need help in the category of people who did not try hard enough. Her sisters’ kitchens are, to her, evidence of character. The superiority lives here, and so does her experience of asking for help as a failing grade entered in permanent ink.
Where a child learns this
The pattern was visible before the two experiences she later identified as formative. Her kindergarten teacher wrote home that Caroline had recopied a full page because one letter touched the line, and a second-grade report card, read during the record review, says that she is a pleasure to teach and needs to learn that mistakes are part of learning. Her father, a mechanical engineer, irons five shirts every Sunday and has never been evaluated for anything, there being nobody in his life with the standing to suggest it. An inherited disposition is plausible, since twin research puts the heritability of the stable liability to these traits at about half.12
Two things happened to her, three years apart. When she was nine she sat through a timed arithmetic test needing the restroom, because leaving would have meant an unfinished page, and she wet herself at her desk. A classmate’s nickname for her lasted into middle school. Her mother collected her, drove home in silence, and said at the front door that they would not be telling her father. When she was twelve her father found a 79 on a pre-algebra examination that she had hidden in her backpack. He did not raise his voice, and he retaught her the unit at the kitchen table every night for two weeks while her sisters watched television in the next room.
A nine-year-old learned that a body which escapes control is unspeakable, and a twelve-year-old learned that to be helped is to be displayed as someone who needed it. Research on early shame experiences finds that they carry the properties of traumatic memory,13 and that experiencing shame as unbearable tracks with personality disorder symptoms, obsessive-compulsive ones included.14
Perfectionism in a child like this is a defense that works. If nothing is ever wrong, nobody has grounds to look at her that way again, and for fifteen years the strategy paid, since teachers adored her and the trait was rewarded at every level of a school system built to reward it, which is also how well it hides. The formulation that best organizes her history is that the two experiences hardened a temperament into something closer to a disorder, and that its impairment became unmistakable in adult life, which asks for what the defense forbids, closeness and a tolerance for experiments that fail for no reason. Direct research on early adversity in OCPD is thin,15 so the account rests on her records and her history.
Why it resembles grandiosity
Caroline can sound grandiose. She is certain the lab would run properly if everyone adopted her methods, and her contempt for how her sisters eat is undisguised. More than one clinician had wondered about narcissism, reasonably, since the perfectionism that makes demands of other people is the form tied to narcissistic grandiosity in the research.9
The two conditions overlap, and shame can sit under a show of superiority in either. The manual’s guidance is a matter of likelihood. People with narcissistic personality disorder may profess the same perfectionism, but they are more likely to believe they have achieved it, while people with OCPD are usually self-critical.7 What the evaluation found in Caroline is better described as overcompensation. Her superiority is specific to standards and carries no claim to special status, and she met one of nine narcissistic criteria on structured interview. When the freezer failed she stayed thirty-one hours re-racking dead samples, and in the week that followed she ran farther and ate less.
The superiority is real and she enjoys it. It is also a wall, and the task of the evaluation was to describe what the wall protects.
What the testing showed
Caroline treated the evaluation as an examination. She arrived twenty minutes early to each of four sessions carrying a typed symptom chronology under numbered headings. She asked whether the tests were graded on a curve and what the average was, and after copying a complex figure she asked to do it again. She inquired whether there was a study guide, and there is not, although the instinct was admirable. All of it was recorded, because how a person takes a test is part of the result.
The battery was built for her question and adjusted as the findings came in. The scores below, which are illustrative values constructed for this case, are the ones that carried the argument.16
| Question | Measure | Result | Reading |
| Can the data be trusted? | Performance validity measures; MMPI-3 validity scales | Passed; virtue scales modestly raised | Valid, with a virtuous tilt |
| Is it OCD? | SCID-5 OCD module | Criterion A not met | No obsessions or compulsions; incompleteness probed and absent |
| Y-BOCS, clinician interview | Not rated | No target symptoms qualified; self-report version in her treatment record was 18 | |
| OCI-R | 27 of 72; Ordering 11, Checking 8, Washing 3, Neutralizing 3, Obsessing 2, Hoarding 0 | Positive screen (cutoff 21); Obsessing below its cutoff of 4 | |
| DOCS | 14 of 80; Symmetry 9, Responsibility for harm 4, Contamination 1, Unacceptable thoughts 0 | Below the cutoff of 18 | |
| Is it personality? | SCID-5-PD | OCPD 6 of 8 (threshold 4); Avoidant 2 of 7 (threshold 4); Narcissistic 1 of 9 (threshold 5) | OCPD established |
| MMPI-3 | Compulsivity and Dominance above T 65; Anger Proneness and Disaffiliativeness just below; distress, self-doubt, self-importance, and eating scales within normal limits | Compulsivity and dominance endorsed; little distress endorsed | |
| PID-5, self / sister (0 to 3) | Rigid Perfectionism 2.7 / 2.9; Perseveration 1.1 / 2.0; Intimacy Avoidance 0.8 / 2.5; Restricted Affectivity 1.0 / 2.3; Hostility 0.9 / 2.1; Grandiosity 0.5 / 1.2 | Agreement on perfectionism; accounts diverge on closeness and affect | |
| Hewitt-Flett MPS, raw (15 to 105) | Self-Oriented 96, Other-Oriented 84, Socially Prescribed 71 | One standard for herself and everyone else | |
| Eating and exercise? | EDE-Q 6.0 | Global 2.38; Restraint 4.2, Eating Concern 0.6, Shape Concern 2.5, Weight Concern 2.2 | Restraint about 2 SD above community norms |
| Body mass index | 18.7 | 1.4 pounds above the underweight threshold | |
| How does she think? | WAIS-IV, scaled scores | Verbal subtests 14 to 15; perceptual reasoning 13 to 14; working memory 13; Coding 9, Symbol Search 10 | Verbal reasoning Superior; processing speed Average; no errors on speeded subtests |
| Trail Making Test, Heaton norms | Part A 23 s, scaled score 11, T 43, Below Average; Part B 58 s, scaled score 11, T 39, Mild Impairment; no errors | Slow against demographic peers, without errors | |
| D-KEFS Color-Word Interference | Inhibition scaled score 9; Inhibition/Switching scaled score 8; no errors | Average | |
| Wisconsin Card Sorting Test | Perseverative errors, Below Average | Kept a rule after it stopped working | |
| Rey Complex Figure | Copy 36 of 36 in 7 min 40 s, time at or below the 1st percentile; Immediate Recall T 40, Delayed Recall T 40, Low Average; Recognition T 53, Average | Accurate and very slow copy; recall below her verbal memory | |
| CVLT-3, scaled scores | Trial 5 Correct 13; Long-Delay Free Recall 13; High Average | Verbal memory strong |
The OCD measures disagreed with one another. The OCI-R produced a positive screen, with a total of 27 against a cutoff of 21,17 but eleven of those points came from the three ordering items, and the developers reported that the Ordering subscale did not separate their OCD patients from non-patient controls, while the Obsessing subscale separated them best. Caroline’s Obsessing score was 2 against a cutoff of 4, and the DOCS placed her at 14 against a cutoff of 18.18 None of these numbers confirms or excludes OCD, and the interview had to settle it.
On the structured OCD module she reported no obsessions, and her routines did not meet the definition of compulsions, including on direct questioning about incompleteness. The Yale-Brown scale was therefore not rated, because it grades obsessions and compulsions that have already been identified.19 The self-report version in her treatment record shows what happens when a severity scale is pointed at the wrong target. One of its items gives the highest score to a patient who yields to a compulsion completely and willingly, on the premise that the compulsion is unwanted. Caroline yields to arranging her bench completely and willingly because she wants it arranged, and the scale counted her preference as surrender.
The personality data were clear once they were collected. She met six of the eight criteria for OCPD on structured interview, missing only the two concerned with hoarding and with money, and the pattern was documented across settings and back to early childhood. She also met two avoidant criteria, both about the fear of being shamed or embarrassed. On the MMPI-3 she endorsed compulsivity and dominance and little distress or self-doubt. The EDE-Q placed her dietary restraint about two standard deviations above community norms and her global score within one standard deviation of the mean.20
The most informative comparison in the file is between Caroline’s description of herself and her sister’s description of her on the same instrument. They agreed on rigid perfectionism. On intimacy avoidance Caroline rated herself 0.8 and Meredith rated her 2.5, with a similar gap on restricted affectivity. Self-reports and informant reports of personality pathology agree only modestly,21 and a sister sees a person in one setting, so the gap was read alongside what Meredith could describe concretely, which included three years without a conversation that was not about logistics or advice. The evidence supports the conclusion that Caroline understands her standards far better than she understands their effect on the people near her.
The cognitive results need the most care. Her verbal reasoning subtests fall in the Superior range and her processing speed subtests are Average, as much as two standard deviations lower, with no errors on either. On Trail Making Part B she took 58 seconds without an error, which converts to a scaled score of 11 and, for a woman of her age and education, a demographically corrected T score of 39, in the range the Heaton system labels mild impairment, near the 14th percentile.22 That is a measured weakness in timed performance, and neither her reasoning nor her accuracy cancels it. The clinical explanation is that the examiner watched her confirm each connection before drawing it, and the same cautious pace appeared on every timed task. The cost resembles the one that has produced eleven versions of a thesis chapter.
Two other findings fit the same picture. On the card sorting test she went on sorting by a rule after the feedback had changed, and she informed the examiner that the test had become inconsistent. She copied the complex figure perfectly in seven minutes and forty seconds, a time at or below the first percentile for her age,23 building it from the left edge one detail at a time, and her recall of it was Low Average against High Average memory for a word list. A figure encoded as a string of details leaves little structure to retrieve, a mechanism described in OCD,24 and a small neuropsychological study of OCPD found a related profile of set-shifting errors and slow planning.25
None of the cognitive scores separates OCPD from OCD, and none was needed to make the diagnosis. That came from the structured interviews, which tested her behaviors against the criteria for both disorders, from records showing that the pattern predated the experiences blamed for it, and from collateral supplying the half of the picture Caroline does not report. The testing added a measured account of what her habits of mind cost her, in a form she could not dismiss as somebody’s opinion. The battery was assembled around her question, so that the eating measure went in because of a heart rate in an emergency record and the informant form because her own inventory reported so little. Reconciling those sources is what produced the answer, and the largest review of the assessment literature found that separate methods contribute information the others cannot.26 A battery built for the patient and read across its parts is, in my judgment, the highest level of care this field offers and the best way I know to get a diagnosis right, and I would revise that judgment if fixed batteries read score by score were shown to reach the same conclusions.
Why nobody caught it
The first impression was a reasonable one. The emergency physician excluded a cardiac event, which was the job. The psychiatrist heard arranging and checking from a composed, articulate woman in a short visit and made the diagnosis those words most often indicate, and the therapist applied the standard treatment for it in good faith. Where the process failed was later, when five months of treatment produced no change and the diagnosis was kept while the patient was relabeled as resistant. A stalled treatment does not by itself exclude OCD, but it is a reason to reexamine the diagnosis, and here the reexamination took another year to arrive.
The difficulty is structural. When diagnoses from ordinary clinical evaluation are compared with diagnoses from standardized interviews, agreement is low, with a mean kappa of .27 across 38 studies,27 and the gap is at least as wide for personality disorders. In one outpatient practice, unstructured intake interviews identified borderline personality disorder in 0.4 percent of 500 patients, and a structured interview then found it in 14.4 percent of the next 409.28
A personality disorder is a pattern across years and settings, and establishing one takes longitudinal history, collateral sources, structured interviewing, and several hours with the patient. A therapist or a prescriber can gather all of that, and some do, but most practice settings are neither arranged nor paid for it, and a fifteen-minute medication visit certainly is not. Linking cognitive testing to personality data and symptom scales is a further skill again, and a daunting one. So my position is that it is unfair to ask these practitioners to make diagnoses of this size with the data and the tools their settings give them. I mean that as a point about fairness, since their licenses permit the diagnosis and many make it well. An independent evaluator works under different conditions, with no treatment relationship that a hard conclusion might cost and with time enough to watch how a person takes a test.
Online self-tests mean nothing as diagnosis, and Caroline had taken several that agreed with her first one. Even a properly validated screener can mislead when used in a low-prevalence population. At its recommended cutoff the OCI-R identified 65.6 percent of OCD patients and cleared 63.9 percent of non-patients,17 and with about 1.2 percent of adults having OCD in a given year,29 roughly 2 of every 100 positive screens in the general population would be true. A questionnaire completed alone can count how often she arranges things, and it cannot establish what the arranging is for or how far back the pattern runs.
Medication, exposure, and the word surrender
For OCD the evidence is strong. In the trial cited earlier, exposure alone did as well as exposure combined with clomipramine and better than the drug alone.2 Had Caroline had OCD, this was the right kind of treatment.
For OCPD the evidence is thin, and it is better to say so. A systematic review of medication found two randomized trials in total, with 71 and 24 participants with OCPD, and rated the certainty of their findings as very low.30 Medication remains a reasonable question for her psychiatrist, particularly if depression or anxiety declares itself. Caroline’s refusal rests on the conviction that needing a medication is a moral failure, and that conviction is part of the disorder and belongs in the treatment.
Psychotherapy has more to offer. Cognitive behavioral therapy aimed at perfectionism itself has been tested in fifteen randomized trials, with large effects on concern over mistakes and on clinical perfectionism,31 although a reanalysis could not distinguish its effect on other-oriented perfectionism, the dimension that governs how she treats the people around her, from control.32 Radically open dialectical behavior therapy was designed for overcontrolled patients. In the RefraMED trial of treatment-refractory depression it beat usual care on depression at the end of treatment, and the difference was no longer statistically significant at twelve months.33 Among the 117 participants in that trial who also had OCPD, those who received it showed better psychological flexibility and emotional approach coping at twelve months, with no significant difference in depression or interpersonal functioning.34 No treatment for OCPD has definitive empirical support, a judgment its leading researcher made in 2016 and the newer data have not overturned.35
The practical conclusion is to choose a therapist who knows this literature and to stop measuring her progress on an OCD scale. The goals should be ones that can be observed. For Caroline those would be a thesis chapter submitted while she still considers it imperfect, an experiment handed to a labmate and left alone, a meal she did not prepare, a rest day taken because her physician asked for it, and one conversation with a sister that contains no advice.
What goes wrong in therapy with a patient like this is predictable. She turns the therapy into a course and sets out to excel in it, so that sessions fill with insight delivered in complete paragraphs while nothing changes. She grades the therapist, finds the work imprecise, and ends treatment with a courteous email in numbered points. Or the therapist, who is a person, finds her formidable and begins working for her approval. In a study of thirty patients with avoidant or obsessive-compulsive personality disorder in cognitive therapy, improvement was predicted by the strength of the early alliance and by episodes in which the alliance ruptured and was repaired,36 so friction with a patient like Caroline is nothing to avoid, provided the therapist notices it and repairs it.
Then there is the word. Over two years Caroline was told to surrender by a yoga teacher and later by her therapist, and each time she became angrier than the occasion seemed to warrant. She asked the therapist for an operational definition, and she was not being difficult. She has organized her life around the proposition that safety comes from control, and she was being asked to give up the one thing that has ever worked, in exchange for a state nobody could describe, by a method nobody could specify. Asked in the evaluation what she thought the word meant, she said after a long pause that she assumed it meant losing. A person who does not know what surrender is cannot do it on instruction, and the instruction lands as one more assignment she is failing.
What the evaluation gave her, and what it could not
The evaluation changed several things at once. The exposure therapy stopped, and with it the phrase treatment-resistant, which had been describing a mismatch between treatment and disorder all along. Caroline’s eating and her heart rate went to a physician and an eating disorder specialist for assessment, having previously gone to nobody. Meredith came away with a different understanding of her sister’s distance, and she said it changed how she heard every piece of unsolicited advice Caroline had ever given her.
Caroline took the feedback session the way she had taken the tests. She brought a notebook, asked how she had done, and wanted to know which scores were the bad ones. When the diagnosis was explained she listened to the whole of it and said that the description was accurate and that she did not see the problem, since what had been described was a person with high standards. She asked whether the personality interview could be scored again. She found a typographical error on page fourteen of the report, and she was right about it.
In the weeks that followed she read the report four times and sent three pages of numbered objections, most of them fair on points of detail and none of them touching the conclusion. She bought the workbook that had been recommended and finished it alone in one sitting with a highlighter, which is the opposite of how it was recommended to her. She saw the recommended therapist twice, concluded that he was imprecise, and wrote him a courteous email. She went to her niece’s birthday party with a well-chosen gift, stayed forty minutes, and left before the cake.
An evaluation of this kind can establish a defensible diagnosis and formulation with a good deal of confidence, and it is some of the most useful work the field has to offer. It has no power to make her want what it found. Without that, the most careful report in the practice is a document in a drawer, and the honest summary of this one is that it changed Caroline’s treatment team more than it has changed Caroline.
The last item in the file is an email sent at 5:41 on a Monday morning, thanking the evaluator for a thorough report and attaching a list of two further typographical errors, both of them real. The report itself is in the second drawer of her desk, in a folder labeled in her handwriting, between her lease and Franklin’s vaccination records. It is filed correctly, and as far as anyone knows it has not been opened since.
Caroline is a composite. The details belong to no single patient, assembled because this pattern arrives often, wearing different clothes.
Resources
These are specialty and academic programs, several of them private pay, grouped by the question they answer.
When the diagnosis itself is in question
- Neuro Assessment Center, Encino, California. Comprehensive neuropsychological and psychodiagnostic evaluation with record review and collateral interviews. Private pay. (310) 926-8336.
- The Pavilion at McLean Hospital, Belmont, Massachusetts. A two-week, self-pay residential program devoted to diagnostic evaluation and treatment planning.
For obsessive-compulsive personality disorder
- Northwell Health OCD Center at Zucker Hillside Hospital, 75-59 263rd Street, Glen Oaks, New York. Directed by Anthony Pinto, PhD, and the program most closely identified with OCPD research and treatment, serving patients seventeen and older with OCD or OCPD. Accepts most insurance plans. (718) 470-8052.
- Austen Riggs Center, Stockbridge, Massachusetts. An open, voluntary residential setting for complex and treatment-resistant personality presentations, beginning with a six-week evaluation and treatment period and built on psychotherapy four times a week. Largely self-pay. Admissions 800-517-4447.
- The OCPD Foundation. Clinician training and referral resources specific to OCPD.
For OCD once it has been confirmed
- McLean OCD Institute, Belmont, Massachusetts. Residential and partial hospital treatment for severe OCD.
- UCLA Adult OCD Intensive Outpatient Program, Los Angeles. Daily exposure-based treatment on weekday mornings for up to thirty treatment days. Evaluation line 1 (800) 825-9989, option 2.
- International OCD Foundation. The directory of clinicians and intensive programs with specialty training in OCD.
For eating disorder assessment and treatment
- Klarman Eating Disorders Center at McLean Hospital, Belmont, Massachusetts. Residential and partial hospital treatment for young adults ages 18 to 28, including cisgender women, transgender women, and non-binary individuals, with anorexia, bulimia, binge eating disorder, or ARFID. 617-855-3410.
Two books
- Anthony Pinto and Michael Wheaton, The Obsessive-Compulsive Personality Disorder Workbook (New Harbinger, 2026). Written by two of the clinicians most identified with OCPD research, and best worked through with a therapist, a chapter at a time.
- Jonathan Abramowitz, Getting Over OCD: A 10-Step Workbook for Taking Back Your Life, second edition (Guilford, 2018). The exposure-based program for confirmed OCD, and the wrong book for the patient in this article.
Notes
1. Glazier K, Swing M, McGinn LK. Half of obsessive-compulsive disorder cases misdiagnosed: vignette-based survey of primary care physicians. J Clin Psychiatry. 2015;76(6):e761-e767. Overall misidentification 50.5 percent; symmetry vignette 3.7 percent.
2. Foa EB, Liebowitz MR, Kozak MJ, et al. Randomized, placebo-controlled trial of exposure and ritual prevention, clomipramine, and their combination in the treatment of obsessive-compulsive disorder. Am J Psychiatry. 2005;162(1):151-161.
3. Pinto A, Liebowitz MR, Foa EB, Simpson HB. Obsessive compulsive personality disorder as a predictor of exposure and ritual prevention outcome for obsessive compulsive disorder. Behav Res Ther. 2011;49(8):453-458. The 23 to 32 percent comorbidity range comes from this paper’s review of earlier work.
4. Grant BF, Hasin DS, Stinson FS, et al. Prevalence, correlates, and disability of personality disorders in the United States: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry. 2004;65(7):948-958. Seven of the ten DSM-IV personality disorders were assessed.
5. Pinto A, Steinglass JE, Greene AL, Weber EU, Simpson HB. Capacity to delay reward differentiates obsessive-compulsive disorder and obsessive-compulsive personality disorder. Biol Psychiatry. 2014;75(8):653-659.
6. Summerfeldt LJ. Understanding and treating incompleteness in obsessive-compulsive disorder. J Clin Psychol. 2004;60(11):1155-1168.
7. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision. 2022. Criteria for obsessive-compulsive disorder; criteria and differential diagnosis for obsessive-compulsive personality disorder.
8. Cain NM, Ansell EB, Simpson HB, Pinto A. Interpersonal functioning in obsessive-compulsive personality disorder. J Pers Assess. 2015;97(1):90-99.
9. Smith MM, Sherry SB, Chen S, Saklofske DH, Flett GL, Hewitt PL. Perfectionism and narcissism: a meta-analytic review. J Res Pers. 2016;64:90-101.
10. Anderluh MB, Tchanturia K, Rabe-Hesketh S, Treasure J. Childhood obsessive-compulsive personality traits in adult women with eating disorders: defining a broader eating disorder phenotype. Am J Psychiatry. 2003;160(2):242-247.
11. Young S, Rhodes P, Touyz S, Hay P. The relationship between obsessive-compulsive personality disorder traits, obsessive-compulsive disorder and excessive exercise in patients with anorexia nervosa: a systematic review. J Eat Disord. 2013;1:16.
12. Gjerde LC, Czajkowski N, Røysamb E, et al. A longitudinal, population-based twin study of avoidant and obsessive-compulsive personality disorder traits from early to middle adulthood. Psychol Med. 2015;45(16):3539-3548. Heritability of the stable liability to OCPD traits was 0.53.
13. Matos M, Pinto-Gouveia J. Shame as a traumatic memory. Clin Psychol Psychother. 2010;17(4):299-312.
14. Schoenleber M, Berenbaum H. Aversion and proneness to shame in self- and informant-reported personality disorder symptoms. Personal Disord. 2012;3(3):294-304.
15. Birgenheir DG, Pepper CM. Negative life experiences and the development of cluster C personality disorders: a cognitive perspective. Cogn Behav Ther. 2011;40(3):190-205.
16. Instruments named in the table: Structured Clinical Interview for DSM-5 and its personality disorders version (First and colleagues); Yale-Brown Obsessive Compulsive Scale; Obsessive-Compulsive Inventory-Revised; Dimensional Obsessive-Compulsive Scale; MMPI-3 (Ben-Porath and Tellegen, 2020); Personality Inventory for DSM-5 and its informant form (Krueger and colleagues, 2012; Markon and colleagues, 2013); Hewitt-Flett Multidimensional Perfectionism Scale (1991); Eating Disorder Examination Questionnaire 6.0 (Fairburn and Beglin, 2008); WAIS-IV; Trail Making Test; D-KEFS; Wisconsin Card Sorting Test; Rey Complex Figure Test; California Verbal Learning Test, Third Edition. The scores are illustrative values constructed for a composite case.
17. Foa EB, Huppert JD, Leiberg S, et al. The Obsessive-Compulsive Inventory: development and validation of a short version. Psychol Assess. 2002;14(4):485-496. Development-sample means were 28.0 for OCD patients and 18.8 for non-anxious controls; Ordering 4.76 versus 4.40.
18. Abramowitz JS, Deacon BJ, Olatunji BO, et al. Assessment of obsessive-compulsive symptom dimensions: development and evaluation of the Dimensional Obsessive-Compulsive Scale. Psychol Assess. 2010;22(1):180-198.
19. Goodman WK, Price LH, Rasmussen SA, et al. The Yale-Brown Obsessive Compulsive Scale. I. Development, use, and reliability. Arch Gen Psychiatry. 1989;46(11):1006-1011.
20. Mond JM, Hay PJ, Rodgers B, Owen C. Eating Disorder Examination Questionnaire (EDE-Q): norms for young adult women. Behav Res Ther. 2006;44(1):53-62. Restraint mean 1.30, SD 1.40; global mean 1.52, SD 1.25.
21. Klonsky ED, Oltmanns TF, Turkheimer E. Informant-reports of personality disorder: relation to self-reports and future research directions. Clin Psychol Sci Pract. 2002;9(3):300-311. Median correlation for dimensional scores was .36 and median kappa for categorical diagnoses was .14; for obsessive-compulsive personality disorder specifically, the medians were .35 and .11.
22. Heaton RK, Miller SW, Taylor MJ, Grant I. Revised Comprehensive Norms for an Expanded Halstead-Reitan Battery. Psychological Assessment Resources; 2004. Appendix C for raw to scaled scores; Appendix D cell for women with 18 to 20 years of education, ages 20 to 34.
23. Meyers JE, Meyers KR. Rey Complex Figure Test and Recognition Trial: Professional Manual. Psychological Assessment Resources; 1995. Normative table for ages 25 to 29.
24. Savage CR, Baer L, Keuthen NJ, et al. Organizational strategies mediate nonverbal memory impairment in obsessive-compulsive disorder. Biol Psychiatry. 1999;45(7):905-916.
25. Fineberg NA, Day GA, de Koenigswarter N, et al. The neuropsychology of obsessive-compulsive personality disorder: a new analysis. CNS Spectr. 2015;20(5):490-499.
26. Meyer GJ, Finn SE, Eyde LD, et al. Psychological testing and psychological assessment: a review of evidence and issues. Am Psychol. 2001;56(2):128-165.
27. Rettew DC, Lynch AD, Achenbach TM, Dumenci L, Ivanova MY. Meta-analyses of agreement between diagnoses made from clinical evaluations and standardized diagnostic interviews. Int J Methods Psychiatr Res. 2009;18(3):169-184.
28. Zimmerman M, Mattia JI. Differences between clinical and research practices in diagnosing borderline personality disorder. Am J Psychiatry. 1999;156(10):1570-1574.
29. Ruscio AM, Stein DJ, Chiu WT, Kessler RC. The epidemiology of obsessive-compulsive disorder in the National Comorbidity Survey Replication. Mol Psychiatry. 2010;15(1):53-63. The predictive value in the text is an illustration computed from this prevalence and the sensitivity and specificity in note 17.
30. Gecaite-Stonciene J, Williams T, Lochner C, Hoffman J, Stein DJ. Efficacy and tolerability of pharmacotherapy for obsessive-compulsive personality disorder: a systematic review of randomized controlled trials. Expert Opin Pharmacother. 2022;23(11):1351-1358.
31. Galloway R, Watson H, Greene D, Shafran R, Egan SJ. The efficacy of randomised controlled trials of cognitive behaviour therapy for perfectionism: a systematic review and meta-analysis. Cogn Behav Ther. 2022;51(2):170-184.
32. Smith MM, Hewitt PL, Sherry SB, et al. A meta-analytic test of the efficacy of cognitive behavioural therapy for perfectionism: a replication and extension. Can Psychol. 2023;64(4):355-376.
33. Lynch TR, Hempel RJ, Whalley B, et al. Refractory depression: mechanisms and efficacy of radically open dialectical behaviour therapy (RefraMED). Br J Psychiatry. 2020;216(4):204-212.
34. Steinhoff MF, Baudinet J, Hempel RJ, Tillman R, Lynch TR, Gilbert KE. Obsessive-compulsive personality disorder (OCPD) in radically open dialectical behavior therapy (RO DBT) for treatment-refractory depression (TRD). Personal Disord. 2026;17(2):178-188.
35. Pinto A. Treatment of obsessive-compulsive personality disorder. In: Clinical Handbook of Obsessive-Compulsive and Related Disorders. Springer; 2016:415-429.
36. Strauss JL, Hayes AM, Johnson SL, et al. Early alliance, alliance ruptures, and symptom change in a nonrandomized trial of cognitive therapy for avoidant and obsessive-compulsive personality disorders. J Consult Clin Psychol. 2006;74(2):337-345.



