Consensus on diagnostic criteria for Yin deficiency syndrome in hypertension: a modified Delphi study

HE Yun,HE Jing,XU Xiangru,LI Haixiao,SHI Rongwei,LIANG Junya,ZHOU Ying,ZHU Yao,CHEN Xiaohu,TANG Shuhua,XIAO Min,ZHAN Libin,PEI Yinghao,JIANG Weimin

HE Yun,LI Haixiao,SHI Rongwei,Department of General Internal Medicine,Jiangsu Province Hospital of Chinese Medicine Affiliated to Hospital of Nanjing University of Traditional Chinese Medicine,Nanjing 210029,China

PEI Yinghao,Department of Intensive Care Unit,Jiangsu Province Hospital of Chinese Medicine Affiliated to Hospital of Nanjing University of Traditional Chinese Medicine,Nanjing 210029,China

HE Jing,Department of Hepatology,Jiangsu Province Hospital of Chinese Medicine Affiliated to Hospital of Nanjing University of Traditional Chinese Medicine,Nanjing 210029,China

LIANG Junya,Department of Institute of Hypertension,Jiangsu Province Hospital of Chinese Medicine Affiliated to Hospital of Nanjing University of Traditional Chinese Medicine,Nanjing 210029,China

ZHOU Ying,ZHU Yao,CHEN Xiaohu,TANG Shuhua,JIANG Weiming,Department of Cardiology,Jiangsu Province Hospital of Chinese Medicine Affiliated to Hospital of Nanjing University of Traditional Chinese Medicine,Nanjing 210029,China

XU Xiangru,Department of Emergency,Longhua Hospital affiliated to Shanghai University of Traditional Chinese Medicine,Shanghai 200032,China

XIAO Min,Department of Cardiology,Dongzhimen Hospital Affiliated to Beijing University of Chinese Medicine,Beijing 100700,China

ZHAN Libin,School of Traditional Chinese Medicine &School of Integrated Chinese and Western Medicine,Nanjing University of Chinese Medicine,Nanjing 210023,China

Abstract OBJECTIVE: To reach consensus on the diagnostic criteria for Yin deficiency syndrome in hypertension(YDSH) patients by a modified Delphi method.METHODS: Our study was consistent with T/CACM 1032-2017.The methodology of RAND/UCLA appropriateness was used to develop consensus guidance statements.A nationwide panel of experienced clinical experts from 19 provinces was constructed.These experts were all prominent in Traditional Chinese Medicine (TCM) of cardiovascular diseases.This consensus process consisted of two rounds of questionnaires and a final round of consultation to analyze the weight score of each item.Moreover,the data extraction process is carried out independently by third-party researchers (LIANG Junya,SUN Yang,and DU Xiaona).When there is disagreement in all three rounds,the expert panel group (odd number) are invited to vote,and the one with more votes wins.In the questionnaires,participants were asked to rate the appropriateness of each syndrome item using a nine-point Likert scale.The consensus was defined as a panel median rating 1-3 or 7-9 without disagreement.And then the diagnostic criteria of YDSH were formed according to the weight score in the final round.RESULTS: Twenty-eight experts (84.8%) participated in the first round,and thirty-one (93.9%) finished the second round.After two rounds,the consensus of YDSH was reached on 11 items (25.6%),including symptoms,signs,and pulse condition.Twenty-one experts (63.6%) completed the final round in which they used a grading system for each item.Red tongue with scanty fur had the highest weighting (22.8%),followed by heat in the palms and soles (20.1%).CONCLUSIONS: The consensus-based diagnostic criteria for YDSH,formed by a modified Delphi method,can be widely incorporated in TCM.A further clinical study will be conducted to analyze the diagnosis value and cut-off score of our YDSH criteria.We have registered this study at http://www.guidelinesregistry.org (No.IPGRP-2021CN350).

Keywords: Yin deficiency syndrome;hypertension;consensus;Delphi technique;weight score

1.INTRODUCTION

Hypertension is a progressive cardiovascular syndrome according to literature and clinical research,which mostly belongs to "vertigo" and "headache" in Traditional Chinese Medicine (TCM).1Up to now,it has been agreed that liver-kidneyYindeficiency and ascendant hyperactivity of liverYangare essential of this disease in the field of TCM.Ascendant hyperactivity of liverYang,the pattern ofYindeficiency withYanghyperactivity,and dual deficiency ofYinandYangwere particular TCM syndromes in the early,middle,and late stage of hypertension,respectively.2,3

A study of 7213 cases involving 11 syndrome elements concluded that the syndrome ofYanghyperactivity andYindeficiency were the major two syndromes in hypertension patients.4The current diagnosis of YDSH is usually based on the nature of the disease (Yindeficiency)and the location of the disease (liver and/ or kidney).Althoughtheguidelines for the diagnosis and treatment of hypertension in TCM 5were published by the Chinese Society of TCM in 2011,there was still lack of unified clinical diagnostic criteria for YDSH.As is known to all,the clinical diagnosis of YDSH was subjectively judged by dizziness,tinnitus,soreness and weakness of the lumbus and knees,etc.,according toGuiding Principles for Clinical Research of New Chinese Medicine,6Chinese Internal Medicine7orPractical Chinese Internal Medicine.8Many studies were based on the different diagnosis standards to explore YDSH patients.9-11Regrettably,there is not a unified clinical definition for YDSH.

The diversity of YDSH definitions impeded further investigation and clinical studies.As the lack of diagnostic criteria of YDSH,it is difficult to compare clinical study outcomes and the quality of hospital care in YDSH patients.Consequently,it is an urgent need to identify a widely accepted diagnostic definition of YDSH,enhancing the understanding between clinicians and the comparability of different clinical trials.This study aims to minimize the variation in the diagnostic criteria of YDSH by reaching a consensus on the YDSH definition,with a sub aim to gain more insight into the weight value of syndrome items.The Delphi approach is a remarkably applied consensus method based on experts' views,and will allow us to formulate the diagnostic criteria of YDSH that is eventually supported by a panel of TCM experts.

2.METHODS

2.1.Consensus method

This study used the RAND/UCLA appropriateness method (RAM).12Briefly,the process consisted of two rounds of questionnaires followed by a final round of consultation to analyze the weight score of each item.In the first and second rounds,the ratings are made individually by We-chat to select appropriate rating scores for each item,with no interaction among panelists.In the final round,top-ranking items were selected and performed a weight score evaluation by pairwise comparison among these items.This Delphi study occurred from November 20,2019 to July 27,2020.A summary of the consensus process is shown in Figure 1.

Figure 1 Flow diagram of the consensus process

2.2.Establish an item pool for YDSH

The workflow of the screening in Figure 2 was operated by two independent personnel simultaneously,and a panel of experts agrees upon inconsistencies.Firstly,we conducted a literature search by keywords hypertension andYin-deficiency syndrome in China National Knowledge Infrastructure Database (CNKI),China Science and Technology Journal Database (VIP),Wanfang,and Pubmed.A total of 98 pieces of Chinese literature in the past ten years were found on CNKI by using the search expression "Chinese Library classification number=R2 plus subject=hypertension plus keywords=syndrome".A total of 856 Chinese articles in the past ten years were also detected by using the expression "Classification number=R2 and title or keywords=hypertension,and title or keywords=syndrome differentiation".A total of 133 pieces of Chinese literature were detected with the search expression "subject=hypertension with keywords=Yindeficiency syndrome" in Wanfang database.In PubMed,a total of 66 pieces of Chinese literature were detected with the search expression "subject=hypertension with keywords=Yindeficiency syndrome" finally,we screened a total of 1087 pieces of literature.After comparison and screening,74 papers meet the requirements of the study.Secondly,we reanalyzed these pieces of literature and divided YDSH into different subtypes,including liver-kidneyYindeficiency,kidneyYindeficiency,dual deficiency ofYinandYang,syndrome ofYindeficiency with effulgent fire,the pattern ofYindeficiency withYanghyperactivity,andYindeficiency with hysteria.A total of 1036 valid syndrome items were extracted from the pieces of literature as mentioned above,made into an Excel syndrome items pool and summarized statistically,which includes 51 items according to symptoms,signs,tongue fur,and pulse condition by removing duplicated items (For example,88 times of dizziness,76 times of soreness and weakness of lumbus and knee,72 times of tinnitus,53 times of insomnia,47 times of red tongue,scanty fur 36 times,and so on).Then,after the leader expert panel’s (Chinese Medicine master Zhou Zhongying and famous Chinese medicine professor Tang Shuhua) further revising according to TCM clinical experience,8 items were deleted (forgetfulness,dry mouth and throat,fear of cold and cold limbs,nocturia increased,white fur,chest tightness with chest pain,red tongue with less fluid and humor and Others (less than 2 occurrences)),Among them,red tongue with less fluid and humor was deleted for overlapping with red tongue,another 6 items (fear of cold and cold limbs,nocturia increased,white fur,chest tightness with chest pain,and Others (less than 2 occurrences))were deleted by expert panel discussion according to Internal Medicine of Chinese Medicine and Guidelines for Diagnosis and Treatment of Common Internal Diseases in Chinese Medicine Symptoms in Chinese Medicine(ZYYXH/T4-49.2008),and 1item (dry mouth and throat)was deleted,because which is overlapped with general scale ofYindeficiency syndrome of non-specific diseases in another research group.Finally,a total of 43 items was built as the pool of YDSH syndromes,including symptoms,signs,tongue manifestation,and pulse condition was formed.

Figure 2 Workflow of Screening

2.3.Expert panel

A nationwide experienced clinical panel experts from 19 provinces was constructed.These experts were all prominent in TCM of cardiovascular diseases.All experts were selected from cardiovascular diseases of the Chinese Society of Integrated Traditional Chinese and Western Medicine and China Association of Traditional Chinese Medicine.The number of experts was determined according to the research theme and scale generally 30-50 persons were appropriate.13

2.4.Round 1

The questionnaires in all three rounds were completed by Fa MaiSheng Medical Technology Co.,Ltd.(Nanjing,China),through scan QR codes in We-chat applets.The first round was consisted of 43 items,including symptoms,signs,tongue manifestation,and pulse condition.Participants were asked to select an appropriate rating score for each item using a 9-point Likert scale (0 meant not needed,1 suggested the weakest need,and 9 indicated the strongest need).The end of questions was followed by a text field,where the participants could provide remarks and arguments.We also analyzed experts' self-evaluation of familiarity and basis.

2.5.Round 2

All participants completed the first round received the mean marks and their scores of each item by We-chat,while they did not receive any particular scores of other members.The message also contained the link (QR code)to the second round questionnaire,which consisted of 22 items after the first round screening.The second round's rating rule was identical to the first round.The highest score of tongue manifestation and pulse condition were directly selected for the following questionnaire.Similar symptoms or signs were evaluated by scores and experts’opinion or consensus.Familiarity and basis induced by self-evaluation were also analyzed.

2.6.Round 3

The top 11 ranking items were selected after two rounds of investigations.We performed a weight score evaluation by pairwise comparison among these 11 items in the final round.Therefore,a total of 55 questions,such as Do you prefer red tongue with scanty fur or soreness and weakness of lumbus and knees as items of YDSH,were evaluated by a 9-points scale,which-9 or 9,-7 or 7,-5 or 5,-3 or 3 mean absolute,apparent,slight preference,respectively,for the latter or former and 1 implies no preference.A legend description was built to help experts understand and rate scores accurately.

2.7.Statistical analysis

The consensus was reached if statements were rated appropriate (panel median 7-9 and mean ≥ 6.5) or inappropriate (panel median 0-3 and mean ≤ 2.5) without disagreement.Disagreement was measured by the interpercentile range adjusted for symmetry.This was according to the method used by Moossdorffet al.14MS Excel 2017 (Microsoft Corp,Redmond WA,United States) and SPSS 24.0 (IBM Corp.,Armonk,NY,USA)were applied to conduct the analyses.The final round questionnaire was analyzed by the analytic hierarchy process (AHP).The paired comparison matrix and the consistency check were conducted.If the consistency check is passed,this expertise will be included for TCM symptom weight analysis,and those who failed will not participate in the analysis.The statistical section for this study was completed by the Institute of Hypertension Research,Jiangsu Provincial Hospital of Chinese Medicine.

3.RESULTS

3.1.Expert panel

According to relevant pieces of literature,academic achievement,visibility,and other factors,50 experienced experts were invited to this study.They came from 19 provinces and engaged in the cardiovascular major of TCM for more than 30 years.Twelve experts (24%) did not respond,and five (10%) could not participate due to lack of time.Then a total of the 33-expert panel was built.Twenty-eight (84.8%),thirty-one (93.9%),and twentyone (63.6%) experts completed the first,second,and final round,respectively.The list of panel members,including thirty-two clinicians and one researcher were shown in Table S1.

3.2.Questionnaire round one

As shown in Table 1,the first questionnaire consisted of 43 items.Ten items (23.3%) were rated appropriately without disagreement.Two items (4.7%) were rated inappropriate without dispute.Thirty-one items (72.1%)were voted uncertain in which 12 items were evaluated for the following questionnaire by our leader expert panel (Chinese Medicine master Zhou Zhongying and famous Chinese medicine professor Tang Shuhua).Then,22 out of 43 items (51.2%) were selected for the next round.The mean coefficient of expert authority and item score were 0.88 and 0.36,respectively.Kendall's coefficient of concordance was 0.363 (χ2=426.711,P＜0.001).

3.3.Questionnaire round two

As shown in Table 2,the second questionnaire was an adjusted version of the first round survey.The newly formulated questions contained 22 ranking items.Twelve items (54.5%) were rated appropriate,and ten items (45.5%) were uncertain.Based on expert opinions and voting scores,we selected the item of heat in the palms and soles rather than the item of vexing heat in the chest,palms,and soles.Red tongue and scanty fur and string-like and fine pulse were chosen as the tongue manifestation and pulse condition for consensus,respectively.In conclusion,after two rounds of consensus,it was reached on 12 out of 43 items (27.9%)without disagreement.Then red tongue and scanty fur were merged into one item.Finally,we got 11 items.The mean coefficient of expert authority and item score were 0.89 and 0.25,respectively.Kendall's coefficient of concordance was 0.225 (χ2=97.753,P＜ 0.001).

3.4.Final round

In the final round,we performed a weight score evaluation by pairwise comparison among these 11 items.First,we test the consistency value of 21 expert's questionnaires.The results indicated that eight questionnaires (consistency ratio ranged from 0.101 to 0.481) were eliminated,and 13 consultation forms remained with further weight score analysis.The resultsof the weight score of each item are shown in Table 2.Red tongue with scanty fur had the highest weighting(22.8%),followed by heat in the palms and soles (20.1%).The final diagnostic criteria of YDSH are presented in Table 3.

Table 1 Items in questionnaire round one

Table 2 Items in questionnaire round two

Table 3 Weight score of each item (%)

4.DISCUSSION

The past few years,thanks to a spurt of progress in technology,have seen the standardization and normalization in TCM studies.4,9,15,16YDSH is a kind of pervasive syndrome in hypertension patients.4,9,17,18YDSH patients are major presented withYindeficiency syndrome,such as dry eye,tinnitus,soreness and weakness of the lumbus and knees,etc.11,19,20However,it is still ambiguous about the definition of YDSH and diagnosis weight ofYindeficiency syndrome in YDSH.Consequently,it is a conspicuous work to formulate and revise the standards of YDSH,establishing a standardized theoretical support system for YDSH related scientific research services.In this study,we performed a modified Delphi analysis and reached a consensus on the diagnostic criteria of YDSH after three questionnaire rounds from 33 TCM experts.

The evaluation objects were based on the comprehensive opinions of experts by a quantitative and qualitative prediction-evaluation method,in which expert factors were mainly technical points of application.21In this study,we built an expert panel of 33 domestic cardiovascular TCM experts.The regional and hierarchical differences were calculated and fitted the requirements in Delphi method.The inclusive criteria of experts was based on scholastic achievement and clinical experience,and thus ensured the reliability and authority of this consulting content.The response rates of the first two round questionnaires were 84.8% and 93.9%,respectively,indicated their enthusiasm about this research.We first formed a draft of 43 items after a review of the pieces of literature.Then we performed two rounds of questionnaires using a nine-point Likert scale.The average coefficients of the expert authority were 0.88 and 0.89 in these two rounds,respectively.Meanwhile,Kendall's coefficients of concordance were 0.363 and 0.225 in these two rounds,respectively,indicated a high degree of expert authority,and the follow-up modification plan could be established for the highly coordinated group.After two rounds of results,it reached a more unified recognition,and 11 of 43 items were determined as criteria of YDSH.

In the Guidelines for Clinical Research of New Chinese Medicines (2002 version),6the definition of the pattern ofYindeficiency withYanghyperactivity included five main symptoms and six secondary symptoms.These symptoms were classified into mild,moderate,or severe levels.Unfortunately,these symptoms did not have scoring standards and could not identify the weight value of each item.It is convinced that the grading system promotes the comparing values of hospital care and clinical studies.22Evidence-based practice principles and specific quantitative methods are two main approaches to evaluate and grade TCM syndrome.In this study,we performed a weight score evaluation by pairwise comparison.Red tongue with scanty fur,heat in the palms and soles,and dry eyes had top 3 weight scores with 22.8%,20.1%,and 10.6%,respectively.One study included a total of 104 standard literature that analyzed 104 symptoms characteristics of liver-kidneyYindeficiency syndrome in hypertension.23After screening and comparison,18 symptoms were included in the regression model,among which red tongue with scanty fur (87.5%),heat sensation (59.6%),and dry eyes (25%)are three of the key symptoms of liver-kidneyYindeficiency syndrome in hypertension.There is still lack of reports regarding the value of our top 3 syndromes in YDSH patients.Therefore,we believe that YDSH criteria with weight score would be a better application for clinical and scientific research.

There are some limitations in this study.First,the number of panel members is an inherent limitation of any consensus method.According to the RAM,the panel should consist of a minimum of seven members.13Our panel of 33 experts met the advised panel size.Secondly,due to COVID-19 pandemic,our three-round consultations were performedviathe mobile applet rather than the paper letter that was the initially designed approach.Considering the age of our participants (some experts over 65),they might not be familiar with the process of answering questions on smartphones,especially a large number of questions in the third round.Therefore,11 experts failed to pass the consistency test due to the longer answer time.

Furthermore,the heterogeneous clinical presentation of YDSH remained a challenge for clinical diagnosis.There was not a uniform standard for identifiedYinpulse condition in TCM.Clinicians often recognize each pulse condition by their own experience and comprehension,which would bring reporting bias.The last limitation was that our panel consisted mainly of Chinese medicine clinicians (96.8%),so the homogeneity of our group might cause information bias.Given the clinician as a primary user,we held the idea that our panel composition was appropriate.The diagnostic value of the final 11 items will be evaluated in our further clinical study.

In conclusion,after three rounds of Delphi consultation,we reached the consensus regarding 11 items and their weight scores of YDSH.Moreover,a further clinical study will be conducted to analyze the diagnosis value and cut-off score of our YDSH criteria.

5.ACKNOWLEDGMENTS

We thank Mr.Tao Liyuan,Mr.Yin Qingfeng and Sun Yang,Ms.Du Xiaona for the study assistance.The authors gratefully acknowledge our panel members for their active participation and valuable contribution to this study.