Constitution-Specific Prescriptions and Temperament-Focused Interviewing Within a Single Course of Care

Method — A procedure proposal (protocol proposal) and the design of a first validation study, synthesizing the findings of parts 1–3 of the series. This is not a study that demonstrates efficacy.

• Sources: the classical texts collated in parts 1–3 of the series (the 1901 Sinchuk edition of Dongui Suse Bowon in the Korean Medicine Classics Database, and the original texts of Dongmu Lee Je-ma's collected writings), the clinical chapters of Saeroseun Sasang Uihak (Sasang Medicine, Newly Written), Sasang Simhak (Sasang Psychology) and the records of Taeyul, and PubMed meta-analyses and randomized controlled trials (searched 2026-09-27; checked against abstracts, with the review on cardiac neurosis checked against the full text). No formal quality appraisal was performed.
• Attribution: the text distinguishes throughout between what is taken from the classical texts and clinical records, what is taken from modern research, and what is newly specified by this paper.

1. Summary

Part 1 of this series showed that two clinical bodies of literature attend first to different things: one approach centres on constitution-specific prescriptions, the other on the analysis of temperament (seongjeong, 性情, the patterned dispositions of nature and emotion that Sasang medicine assigns to each constitutional type). Part 2 reconstructed, from the surviving records, the assessment, interviewing, diary analysis and therapist training of Taeyul's Sasang Simhak (Sasang psychology), and Part 3 set out five elements of the intervention hypotheses that the classical text permits once personality pathology has been described in terms of bi-bak-tam-na (鄙薄貪懦; meanness, frivolity, greed and timidity, the four vices named in the classical text). Part 4 brings these together as a procedure delivered in parallel within a single course of care, and specifies the minimum design of the study that will first test it. The target population of the first study is restricted to adult outpatients with mild-to-moderate depressive symptoms. The first single-arm pilot study evaluates feasibility through the recruitment rate, session completion rate, measurement completion rate and safety events; the BDI-II is collected in order to describe the course of symptoms. The BDI-II at week 12 is reserved as the primary outcome of the subsequent comparative study. Temperament-focused interviewing is delivered once a week for 30 minutes, 12 sessions in total, and constitution-specific herbal prescriptions (chejilbang, formulas selected according to the patient's Sasang constitutional type) are given in parallel over the same period. The stages of hangsim (恒心, the habitual or constant state of mind that the classical text attributes to each constitutional type), the relationship map and the four-position code are recorded not as outcome measures but as exploratory indicators under development. Modern research provides a background for studying the combination of medication with psychological intervention, and for a model of care in which measurement results inform treatment decisions; it does not, however, support the effect of the specific combination of constitution-specific prescriptions and temperament-focused interviewing. Validation is proposed in the following order: reliability of assessment, then feasibility, then standardization of case records, then a comparative study.

---

2. Context of the Question

What emerged over the three preceding papers is that the necessary pieces are located in different places. Explicit criteria for prescription choice and concrete management guidance are found in the literature centred on constitution-specific prescriptions. Methods for reading temperament and working with it through interviewing are found in the Sasang Simhak records. Principles concerning stages and relapse, and concerning relationships and boundaries, are scattered through the classical text. Standards for measurement and for collaborative care are found in modern psychiatry.

In the records examined in Parts 1 and 2, temperament did not enter the diagnostic axis on the prescription-centred side, while on the Sasang Simhak side neither the procedure nor its outcomes were recorded in a standardized way. This paper therefore asks two questions. In what form can these pieces be bound together? And, in order to establish whether what has been bound together actually works, how should the first study be designed?

---

3. What the Literature Says: The Procedure and Its Basis

0. Minimum Specification of the First Validation Study

Before describing each step of the procedure, this section fixes the specifications that will hold for the first single-arm pilot study and for the subsequent comparative study. The specifications below were not set by the classical texts or by modern research; they are set by this paper.

Step 1. Safety Screening and Collaborative Care

The clinical chapters of Saeroseun Sasang Uihak recommend hospital admission in mood disorders when there is a history of suicide attempts, a psychotic tendency or severe mania, and note that mood disorders caused by brain injury respond poorly to treatment. Taeyul himself also recorded that analytic treatment carries considerable risk for patients with schizophrenia. The present procedure translates these descriptions into the exclusion criteria listed above. Any adjustment of psychiatric medication the patient is already taking is decided in consultation with the prescribing clinician. Because a qualitative study that interviewed 26 Korean medicine doctors identified insufficient institutional support and fragmentation as barriers to collaboration between conventional and Korean medicine, the referral pathway is to be fixed before the first study begins.

Step 2. Assessment on Two Levels

The outcome level uses validated scales only. A study of Korean medicine treatment for depressive disorder reported in the Journal of Oriental Neuropsychiatry (102 patients, retrospective, single-arm) measured the BDI-II, STAI, BAI and STAXI every 4 weeks, and all scales had decreased significantly after 12 weeks. The decision of the present procedure to measure the BDI-II at 4-week intervals, and to reserve it as the primary outcome of the subsequent comparative study, was made so that results can be compared with that study.

The exploratory level holds the records of temperament. Records at this level are not used to judge outcomes; their purpose is to explore what can be recorded and how. Constitutional diagnosis is documented, as discussed in Part 2, by writing out the grounds for the diagnosis in sentences. The four-position code (a notation that describes a person's pattern across four positions, sa 事 affairs, sim 心 mind, sin 身 body and mul 物 things) is recorded only after its notation rules have been settled, because the rules given in the commentary and the notation used in the 1996 records (with slashes) have not been reconciled with each other. The hangsim record was prompted by the following passage of the classical text.

太陰人, 恒有怯心, … 若怯心, 至於怕心, 則大病, 作而怔忡也.

(The Taeeumin always has a timid mind … if the timid mind reaches the point of a dreading mind, a serious illness arises and becomes palpitation with fright [jeongchung].)

— Dongui Suse Bowon (東醫壽世保元), Discussion on Differentiating the Four Constitutional Types [17-11] (mediclassics.kr/books/182/volume/4#content_236)

The classical text records all three positions (the everyday state of mind, the deepened stage and the resulting disorder) only for the Taeeumin (怯心→怕心→怔忡; geopsim, timidity → pasim, dread → jeongchung, palpitation with fright) and the Soyangin (懼心→恐心→健忘; gusim, apprehension → gongsim, fear → geonmang, forgetfulness). For the Soeumin and the Taeyangin, only the everyday state of mind (不安定之心, an unsettled mind; 急迫之心, a hurried mind) and the outcome when it becomes settled are described. The three-column form "everyday → deepened → disorder", applied in common to all four constitutions, is therefore a reconstruction by this paper and is not yet an assessment instrument. The relationship map (the categories of friendship, factional associates, affairs and dwelling in the Hwakchung [Expansion] chapter, and the supporting persons at each life stage described in the Gwangjeseol [Discourse on Broad Salvation]) is placed on the same level. It is assumed from the outset that, even if raters agree on such records, agreement does not establish validity, that is, that the records actually measure real change.

Labels such as bi-bak-tam-na are not used in the records. As decided in Part 3, the record states, in place of a label, the behaviour observed, the situation in which it occurred, and the uncertainty of the interpretation. The two items that Part 3 proposed in place of the prognostic descriptions of the classical text, namely whether the patient concretely reviews their own behaviour and its consequences, and whether the patient actually attempts the changes agreed upon, are also recorded as exploratory indicators. Repetitive self-blame and living conditions such as livelihood, caregiving and cost are recorded separately, so that they do not contaminate the interpretation of these items.

Step 3. Treatment Contract

The classical text distinguishes the healing of an illness from the making whole of a person.

凡無論某病人, 恭敬其心, 蕩滌慾火, 安靜善心, 一百日, 則其病無不愈, 二百日, 則其人無不完.

(Whatever the illness, if the patient reveres their own mind, washes away the fire of desire and settles the good mind, then in one hundred days there is no illness that is not healed, and in two hundred days there is no person who is not made whole. Translation by the author.)

— Dongui Suse Bowon, Taeeumin [13-25] (mediclassics.kr/books/182/volume/4#content_42)

The present procedure borrows this structure to explain the goals of treatment on two levels. The course of symptoms is recorded with the BDI-II, and change in temperament and relationships is recorded with the exploratory indicators. The day counts of the classical text are not adopted as such; the contract states the week-12 assessment point and the interim measurement points (weeks 4 and 8). The contract also includes the frequency of interviews, whether a guardian attends, the principles governing concomitant medication, the exclusion and referral criteria, and how to make contact in the event of risk.

Step 4. Parallel Intervention with Two Foci

The bodily focus. Following the method of the prescription-centred literature, the prescription is chosen by considering organ-pattern differentiation (jangbu byeonjeung, the identification of disorder patterns according to the functional state of the viscera and bowels) together with the constitutional type. The grounds for the choice of prescription and the pattern differentiation are recorded, so that when another Korean medicine doctor reviews the same record it is possible to see at which point their judgement would diverge. Lifestyle guidance is given by writing down the times for rising, meals and walking and handing them to the patient. A systematic review conducted to develop Korean medicine clinical practice guidelines for cardiac neurosis included 142 papers, counted 151 comparisons as units of analysis, and reported that groups receiving herbal medicine together with psychotherapy had better overall effectiveness than groups receiving conventional medication together with psychotherapy. All included studies, however, were conducted in China, heterogeneity between studies was large, and methodological quality was low. The combination of herbal medicine with psychological intervention has thus been studied, but the evidence is weak and the populations differ from that of the present procedure.

The temperament focus. Interviewing follows the three principles reconstructed in Part 2 (patient, environment, adaptation) and addresses the five elements of the intervention hypotheses presented in Part 3 (entry, target, direction, relationship, time). Because Part 3 presented these elements only as items for consideration, the present procedure proposes an order and transition conditions as follows.

Under the direction element, the four directions of the Dokhaeng (Solitary Conduct) chapter are not merged into one. Geun (謹, prudence: conducting oneself with restraint and care), toe (退, stepping back: withdrawing from pushing oneself forward or pressing ahead), wan (緩, slowing haste: easing one's impatience) and sil (實, diligent substance: practising steadily and building real substance) must each be translated into a distinct behavioural task, and how each direction is to be made concrete as a task is a question to be worked out from the case records of the first study. The order and conditions above do not appear in the classical text or in the Sasang Simhak records; they are proposals of this paper and will be adjusted in the feasibility study.

The diary uses a three-column form that this paper reconstructed from the record of one case in Part 2 ("Diary analysis: problem / problem analysis / points for ego strengthening"). When the patient brings brief notes of how they felt each day, the therapist records, in separate columns, the observed statements, the therapist's interpretation and the task agreed together. Interpretations are marked as hypotheses rather than facts, and alternative interpretations are recorded alongside them. Tasks are not imposed unilaterally by the therapist. When proposing a task, the therapist explains its purpose, records whether the patient agreed and how they responded, and at the next session first asks whether the patient tried the task and what the experience was like. Whether to convey an interpretation to the patient, and in what words, is likewise decided in this process by observing the patient's response, and that judgement is recorded.

Regarding the combination of medication with psychological intervention, modern research is consulted only as background. In a meta-analysis of combined treatment for major depressive disorder (23 trials, 2,184 participants), the combination of psychotherapy and antidepressants was superior to antidepressants alone when followed up for 6 months or more after randomization (odds ratio 2.93, 95% CI 2.15–3.99), but did not differ from psychotherapy alone in long-term response. The effect of adding active medication to psychotherapy was smaller than, but significantly different from, adding placebo (standardized mean difference 0.25). Both studies concern antidepressants; they can serve only as background for studying an integrated approach and do not support the effect of the specific combination of constitution-specific prescriptions and temperament-focused interviewing.

Step 5. Sharing of Measurements, Scheduled Review and Safety Response

Scheduled review (provisional criteria). The present procedure measures the BDI-II and the secondary scales at weeks 4, 8 and 12, reviews the results together with the patient, and reflects them in the prescription and the interview plan. The first review of response takes place at week 4. If the patient has worsened at week 4, or has shown no improvement by week 8, the prescription and interview plan are re-examined and referral is discussed. These criteria are not established rules but provisional proposals of this paper, to be adjusted in the feasibility study. A meta-analysis of measurement-based care for depressive disorders (7 trials, 2,019 participants) reported higher remission rates than usual care (odds ratio 1.83, 95% CI 1.12–2.97; the difference in response rates was not significant). All included studies, however, involved pharmacotherapy, and what this study supports is the process of reviewing measurement results with the patient and reflecting them in treatment decisions, not the specific intervals or decision criteria of the present procedure. In a 24-week randomized trial (154 participants), the measurement-based care group likewise reached response and remission sooner, but there was no significant between-group difference in response and remission rates at week 24.

Safety response (independent of scheduled dates). Separately from the scheduled reviews, suicidal ideation, self-harm, rapid deterioration and adverse effects are checked at every session. If any of these occur, assessment is carried out on the spot without waiting for the scheduled assessment date, and if there is immediate risk, urgent action is taken, such as same-day emergency referral to psychiatry. If an adverse effect of the herbal medicine is suspected, the prescription concerned is stopped immediately and the event is recorded.

Relapse and assessment points. For haematemesis in the Soyangin, the classical text states that if it recurs, "counting from the day of the recurrence, in one hundred days it will be somewhat better" (Dongui Suse Bowon [11-3]). The present procedure takes from this passage only the structure of re-evaluation. If worsening or relapse occurs during the study period, it is recorded as a separate event and the patient's treatment plan is revised, but the scheduled assessment points of the study (weeks 4, 8 and 12, and the follow-up point) are not moved. If assessment points were postponed with each event, comparison between participants would become impossible.

Requirements on the Therapist's Side

Sasang Simhak presupposes 5–10 years of self-analytic training and dual supervision (Part 2). The present procedure replaces this with a monthly scheduled case review and a session-record checklist. This replacement is not found in the Sasang Simhak literature; it is a change newly proposed by this paper, and it is unknown to what extent it can substitute for the correction of observer bias that the long-term training was intended to achieve. That the relationship between therapeutic alliance and outcome was reported to be of moderate size across 79 studies is a reason to monitor the therapeutic relationship, but it does not support the effectiveness of case review as a method.

---

4. Cross-Reading: Summary of the Procedure and the Source of Each Element

The very length of the last column shows the character of this procedure. The classical texts and records indicate what should be observed, and modern research provides the principle that measurement should inform treatment decisions; but most of the concrete specifications and criteria have been set by this paper and are themselves the object of validation.

---

5. What Is Not Yet Known: The Order of Validation

First, reliability of assessment. After the notation rules of the four-position code have been finalized, two or more therapists independently rate the constitutional diagnosis, the code and the hangsim record for the same patients, and inter-rater agreement is examined. Even if agreement is confirmed, it constitutes reliability; the validity of these records as measures of real change must be examined separately.

Second, feasibility. A single-arm pilot study applies the procedure, under the minimum specification above, to a small number of patients. It examines the recruitment rate, 12-session completion rate, diary submission rate, task attempt rate, scale measurement completion rate and safety events. This is not a stage at which effectiveness is judged, and the BDI-II is collected solely to describe the course of symptoms.

Third, standardization of case records. Taeyul's structured online case record examined in Part 2 contained a narrative of short-term improvement, but it ended as an interim record one month after the first interview, without standardized assessment. The present procedure aims at a case series in which the scales and the exploratory indicators at baseline, week 4, week 8, week 12 and follow-up are kept in the same format.

Fourth, a comparative study. Once the above stages are in place, constitution-specific prescription alone can be compared with the combined procedure, and at this stage the BDI-II at week 12 becomes the primary outcome. The question this comparison answers is the add-on effect of the whole combined procedure, in which interviewing, the diary, tasks and additional contact are bundled together; it cannot separate the respective contributions of the constitution-specific prescription and the interviewing, nor the effect specific to temperament analysis. Scale measurement, sharing of results, the manner of prescription adjustment and the conditions for concomitant treatment are to be kept the same in both arms. The specific effect of temperament-focused interviewing must be asked separately later, with a control intervention matched for contact time and frequency.

The remaining questions are as follows.

1. Can the hangsim record items be used in common for all four constitutions, or should one begin with the constitutions for which the classical text records all three positions?
2. When several therapists use the three-column diary form, how closely do their records of interpretations and tasks agree?
3. How are the four directions of geun, toe, wan and sil made concrete as actual tasks?
4. What relationship do changes in the exploratory indicators show with changes in the BDI-II?
5. Does the monthly case review affect procedural adherence and measures of the therapeutic alliance?

---

6. Source Citation Cards

"Text verified" on each card indicates the extent to which the source was actually checked in the series. Because no formal quality appraisal was performed, no certainty-of-evidence grade has been assigned. Detailed cards for the classical and school-specific literature are given in parts 1–3 of the series.

Source 1 [KM]

• Source: Dongui Suse Bowon (東醫壽世保元), 1901 Sinchuk edition, Discussion on Differentiating the Four Constitutional Types [17-11]·[17-12], Taeeumin [13-25], Soyangin [11-3]
• Author/Era: Lee Je-ma, Joseon dynasty (1901 Sinchuk edition)
• MEDICLASSICS: mediclassics.kr/books/182/volume/4#content_236 · #content_237 · #content_42 · volume/3#content_81
• Text verified: collated directly against the original text | Study type: classical text
• Key point: The pathway by which hangsim progresses to a disorder (Taeeumin and Soyangin); the distinction between the healing of an illness and the making whole of a person; recalculation upon relapse (the present procedure borrows only the structure of re-evaluation).

Source 2 [KM]

• Source: Saeroseun Sasang Uihak (Sasang Medicine, Newly Written), clinical chapters, Neuropsychiatry / Sasang Simhak (Sasang Psychology) / Taeyul's records / Gyeokchigo and Dongui Sasang Chobongwon
• Author/Year: Ryu Ju-yeol 2013 / Kang Yong-hyeok 2010 / Kim Do-sun 1990s / Lee Je-ma
• Text verified: see parts 1–3 of the series | Study type: clinical textbook, school commentary, records, classical texts
• Key point: Admission criteria and management guidance; assessment, interviewing and diary records; directions of correction and their boundaries.

Source 3 [WM]

• Source: Combining pharmacotherapy and psychotherapy or monotherapy for major depression? A meta-analysis on the long-term effects
• Author/Year: Karyotaki E et al., 2016, J Affect Disord (PMID 26826534)
• Text verified: abstract | Study type: meta-analysis
• Key point: 23 RCTs, 2,184 participants, ≥6 months after randomization; combined treatment vs antidepressant alone OR 2.93 (2.15–3.99); combined treatment vs psychotherapy alone showed no difference.

Source 4 [WM]

• Source: The contribution of active medication to combined treatments of psychotherapy and pharmacotherapy for adult depression: a meta-analysis
• Author/Year: Cuijpers P et al., 2010, Acta Psychiatr Scand (PMID 19922522)
• Text verified: abstract | Study type: meta-analysis
• Key point: 16 RCTs, 852 participants; effect of adding active medication SMD 0.25 (0.03–0.46).

Source 5 [WM]

• Source: The Efficacy of Measurement-Based Care for Depressive Disorders: Systematic Review and Meta-Analysis of Randomized Controlled Trials
• Author/Year: Zhu M et al., 2021, J Clin Psychiatry (PMID 34587377)
• Text verified: abstract | Study type: systematic review and meta-analysis
• Key point: 7 RCTs, 2,019 participants (all involving pharmacotherapy); remission OR 1.83 (1.12–2.97); response OR 1.66 (not significant).

Source 6 [WM]

• Source: Measurement-Based Care to Enhance Antidepressant Treatment Outcomes in Major Depressive Disorder: A Randomized Clinical Trial
• Author/Year: Husain MI et al., 2025, JAMA Netw Open (PMID 40892412)
• Text verified: abstract | Study type: RCT
• Key point: 154 participants, 24 weeks; shorter time to response and remission; differences in response and remission rates at week 24 not significant.

Source 7 [WM]

• Source: Systematic review and meta-analysis of cardiac neurosis for development of clinical practice guidelines of Korean medicine
• Author/Year: Park HY et al., 2024, Front Psychiatry (PMID 38410677)
• Text verified: full text (Europe PMC PMC10895026) | Study type: systematic review and meta-analysis
• Key point: 142 papers included, 151 comparisons counted as units of analysis, all included studies conducted in China; herbal medicine plus psychotherapy superior to conventional medication plus psychotherapy; large heterogeneity and low quality.

Source 8 [WM]

• Source: Clinical Effects of Korean Medical Treatment on Depressive Disorder using Depression and Anxiety Scales
• Author/Year: An Y (Yunyoung An) et al., 2022, Journal of Oriental Neuropsychiatry 33(3):317–327
• Text verified: abstract | Study type: retrospective single-arm study
• Key point: 102 patients; scales measured at 4-week intervals; significant decreases in BDI-II, STAI, BAI and STAXI after 12 weeks.

Source 9 [WM]

• Source: Exploring collaborative practice between conventional and Korean medicine: a qualitative study of Korean medicine doctors' experienced barriers and strategies
• Author/Year: Park D et al., 2025, BMC Complement Med Ther (PMID 41184826)
• Text verified: abstract | Study type: qualitative study
• Key point: 26 Korean medicine doctors; barriers to collaboration were insufficient institutional support, indifference and fragmentation.

Source 10 [WM]

• Source: Relation of the therapeutic alliance with outcome and other variables: a meta-analytic review
• Author/Year: Martin DJ et al., 2000, J Consult Clin Psychol (PMID 10883561)
• Text verified: abstract | Study type: meta-analysis
• Key point: 79 studies; the relationship between alliance and outcome was moderate in size and consistent.

Source 11 [Author's previous papers]

• Source: Sasang Simhak clinical series, parts 1–3
• Author/Year: Choi Jang-hyuk, 2026, Sowonjae (forthcoming)
• Study type: narrative literature review, methodological reconstruction, interpretation of the classical text
• Key point: The foci of the two approaches; reconstruction of Taeyul's method; intervention hypotheses following bi-bak-tam-na.

---

Choi Jang-hyuk | Doctor of Korean Medicine · Director, Dongjedang Korean Medicine Clinic · Method: procedure proposal and design of a first validation study, synthesizing parts 1–3 of the series