Content Validity of Visual Analog Scales to Assess Symptom Severity of Acute Angioedema Attacks in Adults with Hereditary Angioedema

Content Validity of Visual Analog Scalesto Assess Symptom Severity of AcuteAngioedema Attacks in Adults withHereditary AngioedemaAn Interview Study

Carolyn Vivienne McMillan,1 Jane Speight,1,2,3 Anurag Relan,4 Luca Bellizzi,5 Gerald Haase6

and Marco Cicardi7

1 AHP Research, Hornchurch, UK

2 The Australian Centre for Behavioural Research in Diabetes, Melbourne, VIC, Australia

3 The Centre for Mental Health and Wellbeing Research, Deakin University, Burwood, VIC, Australia

4 Pharming Group NV, Jersey City, NJ, USA

5 Sintesi Research S.r.l., Milan, Italy

6 Gordon & Levine Consultancy Services Ltd, Kingston upon Thames, UK

7 Dipartimento di Scienze Cliniche ‘‘Luigi Sacco,’’ Universita degli Studi di Milano, Ospedale Luigi Sacco,

Milan, Italy

Abstract Background:Hereditary angioedema (HAE) is a rare, debilitating, potentially

life-threatening condition characterized by recurrent acute attacks of edema

of the skin, face/upper airway, and gastrointestinal and urogenital tracts.

During a laryngeal attack, people with HAE may be at risk of suffocation,

while other attacks are often associated with intense pain, disfigurement,

disability, and/or vomiting. The intensity of some symptoms is known only to

the person experiencing them. Thus, interview studies are needed to explore

such experience and patient-reported outcome measures (PROMs) are re-

quired for systematic assessment of symptoms in the clinical setting and in

clinical trials of treatments for acute HAE attacks.

Objective: The aim of this interview study was to assess the content validity

and suitability of four visual analog scale (VAS) instruments for use in clinical

studies. The VAS instruments were designed to assess symptoms at abdominal,

oro-facial-pharyngeal-laryngeal, peripheral, and urogenital attack locations.

This is the first known study to report qualitative data about the patient’s

experience of the rare disorder, HAE.

Methods: Semi-structured exploratory and cognitive debriefing interviews

were conducted with 27 adults with a confirmed clinical/laboratory diagnosisof HAE (baseline plasma level of functional plasma protein C1 esterase in-

hibitor [C1INH] <50% of normal without evidence for acquired angioedema).

ORIGINAL RESEARCH ARTICLEPatient 2012; 5 (2): 113-126

1178-1653/12/0002-0113/$49.95/0

Adis ª 2012 Springer International Publishing AG. All rights reserved.

There were 17 participants from the US and 10 from Italy, with mean age 42.5

(SD 14.5) years, range 18–72 years, mean HAE duration 21.3 (SD 14.1) years,

range 1–45 years, 67% female, and 44% VAS-naıve. Experience of acute

angioedema attacks was first explored, noting spontaneous mentions by par-

ticipants of HAE symptomatology. Cognitive debriefing of the VAS instruments

was undertaken to assess the suitability, comprehensibility, and relevance of

the VAS items. Asymptomatic participants completed the VAS instruments

relevant to their angioedema experience, reporting as if they were experi-

encing an acute angioedema attack at the time. Interviews were conducted in

the clinic setting in the US and Italy over an 8-month period.

Results: Participants mentioned spontaneously almost all aspects of acute

angioedema attacks covered by the four VAS instruments, thus providing

strong support for inclusion of nearly all VAS items, with no important

symptomsmissing. Predominant symptoms found to be associated with acute

angioedema attacks were edema and pain, and there was evidence of varying

degrees of disruption to everyday activities supporting the inclusion of

an overall severity item reflecting the disabling effects of HAE symptoms.

VAS item wording was understood by participants.

Conclusion: This interview study explored and reported the patient experience

of HAE attacks. It demonstrated the content validity of the four anatomical

locationHAEVAS instruments and their suitability for use in clinical trials of

recombinant human C1INH (rhC1INH) treatment for ascertaining trial

participants’ assessments of the severity of acute angioedema symptoms.

Key points for decision makers

� In this interview study, participants spoke of the often severe and distressing symptomsassociated with a hereditary angioedema (HAE) attack, including (depending on the attacklocation) pain, swelling and associated disfigurement, vomiting, and difficulty breathing, andtheir debilitating impact on everyday activities

� There was strong support for inclusion of nearly all the visual analog scale (VAS) items in thefour attack location patient-completed instruments, with no important symptoms missing;VAS items were comprehensible to participants

� The four VAS instruments have good content validity and are considered suitable for use inclinical trials of recombinant human C1 esterase inhibitor (rhC1INH) treatment for ascertainingparticipants’ assessments of the severity of acute angioedema symptoms

Background

Hereditary angioedema (HAE) is an autosomaldominant disease resulting from a heterozygousdeficiency of the plasma protein C1 esterase in-

hibitor (C1INH) caused by mutations of the af-fected C1INH gene.[1,2] It is a rare but seriousdisease with an estimated prevalence of 1 : 50 000worldwide.[3] HAE is characterized by recurrentacute attacks of edema (‘angioedema attacks’),

114 McMillan et al.

Adis ª 2012 Springer International Publishing AG. All rights reserved. Patient 2012; 5 (2)

which mainly manifest themselves in the gas-trointestinal tract (abdominal attacks), the face,the limbs/extremities (peripheral attacks), andmore rarely in the oro-pharynx and larynx (oro-pharyngeal/laryngeal attacks) and urogenital tract.[1]As anatomical demarcation of attacks is difficultin the oro-facial-pharyngeal-laryngeal (OFPL)area, these attacks are often grouped together.For each location, a range of symptoms maybe secondary to edema (intense pain, vomiting,and, in the case of laryngeal edema, the risk ofsuffocation).[1]

The frequency of acute angioedema attacks var-ies widely from very few attacks to almost weeklyoccurrence.[1] Some attacks are triggered (e.g. byphysical trauma, psychological stress, infections,and estrogen), others occur without any noticeabletriggers.[4] Attacks can manifest at one anatomicallocation or simultaneously at several locations, re-sulting in a wide range of presenting clinical symp-toms, which can last from 2 to 5 days beforesubsiding spontaneously.[1] Attacks can impact neg-atively on physical and mental health, daily activ-ities, education, career, and work productivity.[5]

There is no established objective clinical/laboratory assessment to evaluate symptom in-tensity in acute angioedema attacks, thus patient-reported outcome measures (PROMs) must beused to evaluate intensity of some symptoms (e.g.nausea or pain). Visual analog scales (VASs) areamongst the simplest PROMs to complete – asingle 100mm line with descriptive anchors ateach end.[6] They are the standard measurementtool in pain research[7,8] and clinical practice.They are short, practical, and sensitive to treatmenteffects.[9] Thus, VASs are likely to be suitable forcompletion by patients experiencing an acute angio-edema attack and who may be in severe pain. Re-cently, another HAE study has reported successfuluse of VASs.[10]

Condition-specific PROMs are generally moresensitive and responsive to treatment changesthan generic PROMs.[11] HAE-specific VASshave been designed for completion by patients inexploratory open-label studies for recombinant

human C1INH (rhC1INH) treatment, and im-proved for implementation in randomized con-trolled North American and European studies(Pharming Technologies BV).[12] When these stud-ies were designed, no HAE-specific PROMs wereavailable, although a measure of symptom severityand change has been published recently.[13]

The current interview study was undertaken toassess the content validity and suitability of theVAS instruments for use in rhC1INH clinicalstudies. Content validity is an indication of thedegree to which the concept of interest is compre-hensively sampled by the items in the instrument.[14]

It is based on expert judgment, determined on thebasis of (i) exploratory interviews with the targetgroup designed to generate spontaneous mentions1

of relevant issues, and (ii) structured interviews.This process is increasingly referred to as ‘cognitivedebriefing,’[15,16] an interview method during whichparticipants ‘think aloud’ about the comprehensi-bility, relevance, and comprehensiveness of itemsand response options, in reply to standardizedquestioning by the interviewer. It is an essential andaccepted approach to assessing the suitability ofnewly designed PROMs.[17]

Methods

Design and Setting

Semi-structured interviews were used:1. to explore participants’ experience of acuteattacks of angioedema; and2. for cognitive debriefing of participants’ re-sponses to VAS questionnaires with reference toa typical acute angioedema attack for which theywould seek treatment.

Interviews were conducted in four centers(Atlanta [GA], Tampa [FL], and Dallas [TX] inthe US and Milan in Italy) from September 2009to April 2010. The US and Italian study protocolswere approved by theWestern Institutional ReviewBoard and the Ethics Committee of the OspedaleLuigi Sacco, Azienda Ospedaliera, Milan, Italy,respectively.

1 Where participants speak spontaneously about a subject (e.g. use a particular term to describe a symptom)without prior prompting.

Content Validity of Hereditary Angioedema VASs 115


Participants

Approximately equal numbers of adult menand women with a previously confirmed diag-nosis of HAE were to be recruited, with experi-ence of HAE attacks at different anatomicallocations, some of whom would be naıve to VASinstruments.2 Eligible participants met the in-clusion criteria used in the C1-1205 and C1-1304clinical trials (Pharming Technologies BV), i.e.clinical/laboratory diagnosis of HAE with baselineplasma level of functional C1INH <50% of normalwithout evidence for acquired angioedema[12] andalso native speaker of English (US) or Italian (Italy),and age range 18–70 years. Participants were ex-cluded if they were experiencing an acute angio-edema attack at the time of the interview.

Materials

The interview schedule was arranged in twoparts:� Part 1: Semi-structured exploratory interview

designed to investigate HAE symptomatologyfrom the patient’s perspective, to determinesymptom salience and the participant’s choiceof wording.

� Part 2: Cognitive debriefing of the HAEVASs.The participant completed the VAS instru-ment for each symptom location relevant tohis/her ownHAE experience, and was asked tocomment freely before any specific promptingby the interviewer. His/her understanding ofVAS items and opinion on item suitability wasnoted, along with comments about item redun-dancy, or any aspects missing from the VASs.During Part 2, participants completed VAS

instruments for four attack locations: abdominal,

OFPL, peripheral, and urogenital.3 The VAS in-struments were designed for use in clinical trials(in the US and Italy) to assess rhC1INH treat-ment efficacy in relation to changes in the severityand duration of acute angioedema attack symp-toms.[12] Each VAS instrument enables re-spondents to rate the intensity of the followingsymptoms: illness, pain, bloatedness, nausea, andhunger (Abdominal VAS instrument); illness,pain, edema, breathing, speech, and swallowing(OFPL VAS instrument); edema, pain, and use ofextremity (Peripheral VAS instrument); and ill-ness, pain, edema, nausea, urination, and hunger(Urogenital VAS instrument). In addition, foreach location VAS instrument, there is an ‘overallseverity’ VAS to enable a global assessment of theattack intensity at that location. The wording ofthis item is identical for each of the four VASinstruments except the Italian Peripheral VASinstrument, which has a different response optionto the other three Italian location VAS instru-ments.4 Table I shows item wording and abbre-viations.

Each participant also completed a short ques-tionnaire requesting minimal clinical and demo-graphic details. All interview materials for use inItaly were translations of the US materials andproduced by a professional translation agencyusing standard methodology.5 Interview materi-als are presented in the SDC.

Study Administration

One-to-one semi-structured interviews wereconducted by fully qualified psychologists ex-perienced in interviewing techniques (all nativelanguage speakers). Each interview was digitallyrecorded (with participants’ permission) and ver-

2 HAE is a rare condition and recruiting patients who had not previously participated in any HAE research ornot completed a VAS instrument would prove to be very difficult.

3 The Appendix, available as Supplemental Digital Content (SDC), http://links.adisonline.com/PBZ/A36,contains the full US and Italian VAS instruments.

4 Nessun danneggiato/estremamente disabilitanti (Peripheral VAS instrument) versus Nessun sintomo/estremamente disabilitanti (Abdominal, OFPL, and Urogenital VAS instruments).

5 Forward translation into Italian by a native Italian speaker, followed by two back translations into English(performed by certified translators specializing in medical translations who were fluent in Italian and whosemother tongue was English); checks for discrepancies; and then production of final versions.

116 McMillan et al.


batim transcriptions were produced in English.All participants were assured of confidentiality,provided informed consent, and were offered re-imbursement of out-of-pocket expenses for studyparticipation.

Data Analysis

Data from the exploratory phase of the inter-view (Part 1) were subjected to content and the-matic analyses, in accordance with standardmethodologies[18] to identify themes and catego-ries that centered on particular phrases, incidents,or behavior. The numbers of spontaneous men-tions of aspects of symptoms covered in the VASinstruments were noted to determine support forthe inclusion of each VAS item for each attacklocation. Participants did not see the VAS in-struments until Part 2 of the interview, imme-diately prior to cognitive debriefing. This servedto check the degree to which what was saidspontaneously by patients matched the items6 inthe VAS instruments.

Results

Participant Sample

Twenty-seven interviews were conducted: 17 inthe US (Atlanta [9], Tampa [4], Dallas [4]), and 10in Italy (Milan), with a representative sample ofadults with HAE (67% women, 70% Caucasian).7

Respondent characteristics are detailed in table II.

Table I. Visual analog scale item abbreviations used within the text

to summarize the item and response option wording in the hereditary

angioedema instruments

Abbreviation VAS item wording (and response options)

Bloated Do you feel bloated?a

(No, totally empty – Unbearably full)

Breathing How easy is it to breathe?

(No problems breathing – Very difficult)

Desire to eat How would you define your desire to eat?b

(Very weak – Very strong)

Hunger How hungry are you?

(Not at all hungry – Extremely hungry)

Illness How ill do you feel?

(Not at all ill – Extremely ill)

Nausea Do you feel nauseous?

(No nausea at all – Extremely nauseous)

Overall

severity

How severe are the angioedema symptoms now for

this location?

(No symptoms at all – Extremely disabling)

Pain Do you have (abdominal) pain?

(No pain – Extremely painful)

Speech Is your speech affected?

(Not at all – Extremely affected/Cannot speak at all)

Swallowing Do you have difficulty in swallowing?

(Not at all – Great difficulty)

Swelling Do you have any swelling?

(No swelling – Extreme swelling)

Urination How easily can you urinate?

(Easily [no problem] – Cannot urinate)

Using

extremity

Do you have difficulty using the swollen extremity?

(No problems. Can easily move – Extremely

difficult)

a Item is in US Abdominal VAS instrument only.

b Item is in Italian Abdominal VAS instrument only.

VAS = visual analog scale.

6 Reference to VAS items includes their respective response options unless otherwise stated.

7 The US sample of 17 patients was representative in terms of socio-demographic characteristics given the rarityof the HAE condition. Age range was 18–67 years. Thirteen participants (76.5%) were women; four (23.5%) weremen. The majority of participants (n = 13; 76%) were Caucasian. Two (12%) were Black, one person was of Asianorigin, and one other described himself as ‘White (Mexican/German).’ The majority of participants (n = 9; 52.9%)had the equivalent of college education or less but a further six participants (47%) had at least university-leveleducation (Bachelor’s or advanced degree). Ten (59%) were employed, three (18%) were students in full-timehigher education, one was self-employed, one was unemployed, one was a homemaker/carer, and one was retired.The overall preponderance of women in the US sample was not unusual in HAE studies:[19] women tend toexperience more acute angioedema attacks, perhaps due to hormonal factors.[20]

The Italian sample of ten patients was representative given the small size of the sample and rarity of thecondition. Equal numbers of males and females were enrolled, with a range in age (33–72 years), education,cultural, and occupational backgrounds (with five not employed), and with variation in ethnicity (Caucasian, 60%and other unspecified, 40%).



The mean (SD) age was 42.5 (14.5) years, andmean (SD) duration since diagnosis of HAE was21.3 (14.1) years. A family history of HAE wasreported by 74% of participants.

All participants reported having experiencedacute attacks inmore than one location, and almostall in peripheral and/or abdominal locations(table II) with urogenital attacks experienced leastfrequently (63%).8 Fifteen (56%) had completed aVAS previously (six participants in the preceding6 months), and 12 were VAS-naıve. Checks con-ducted on US Part 1 data indicated that satu-ration had been achieved after 12 interviews (fourfrom each US clinic) with little new informationforthcoming thereafter. Thus, the total of 27 in-terviews is considered to be highly satisfactory,particularly given the rarity of HAE.

Part 1: Exploratory Interviews

There is insufficient space here to report thelarge amounts of qualitative interview data gen-erated. See figure S1 in the SDC for some quo-tations concerning participants’ experiences ofHAE. However, content analysis of spontaneous

mentions is reported (see table III), indicating therelevance assigned to each variable by partic-ipants for a given attack location, and hence thecontent validity of the VAS measures.

Terminology Used in theVisual Analog Scales (VASs)

All medical terms in the VASs (e.g. angioedema)were generally well understood by participants, al-though they mostly would not use these termsspontaneously. The term ‘ill’ was not generally usedspontaneously in relation to HAE attacks. Whenparticipants were later asked directly if they felt illduring anHAE attack, eight (47%) US participantssaid they would use ‘ill,’ (five of whom only forabdominal attacks) and nine said they would neveruse ‘ill’ in this context. Their interpretation of ‘ill’included sickness (n= 9) and virus/contagion [e.g.‘flu’] (n= 5). Two Italians spontaneously used theterm ‘ill’ during Part 1 and, when prompted, an-other seven respondents said they would use it.

Abdominal Location

For participants who had experienced abdo-minal attacks (n = 25), the most bothersomesymptoms were pain (n = 21), vomiting (n = 10),

Table II. Participant characteristics and anatomical hereditary angioedema attack locations

US (n = 17) Italy (n = 10) Total (n= 27)

Mean age (SD) [range], y 37.2 (12.7) [18–67] 51.5 (13.5) [33–72] 42.5 (14.5) [18–72]

Women participants [n (%)] 13 (76.5) 5 (50) 18 (66.7)

Ethnicity [n (%)]

Caucasian 13 (76.5) 6 (60) 19 (70.4)

Black 2 (11.8) 0 2 (7.4)

Asian 1 (5.9) 0 1 (3.7)

Other 1 (5.9) 4 (40) 5 (18.5)

HAE duration (SD) [range], ya 19.5 (15.6) [1–45] 24.5 (11.0) [2–34] 21.3 (14.1) [1–45]

HAE attack locations [n (%)]

Abdominal 15 (88.2) 10 (100) 25 (92.6)

Peripheral 16 (94.1) 10 (100) 26 (96.3)

OFPL 13 (76.5) 9 (90) 22 (81.5)

Urogenital 9 (52.9) 8 (80) 17 (63)

VAS-naıve [n (%)] 6 (35.3) 6 (60) 12 (44.4)

a As reported by health professionals in patients’ medical notes.

HAE =hereditary angioedema; OFPL =oro-facial-pharyngeal-laryngeal; VAS = visual analog scale.

8 There is a higher proportion of people with experience of urogenital attacks in this study than is generallyfound because we wanted to ensure that most participants would be able to discuss attacks in all locations.

118 McMillan et al.


and nausea (n = 5). US participants used termssuch as ‘debilitating’ or ‘excruciating,’ to describepain as symptoms peaked.

Spontaneous mentions: Almost all reportedabdominal edema (n = 24), with severe, frequentto continuous vomiting episodes as the attackpeaked (n = 21), preceded and/or accompanied bynausea (n= 20). Loss of appetite (n= 13), dehy-dration (n= 10), diarrhea (n= 8), light-headedness(n = 7), and fatigue (n = 6) might occur. Some USparticipants specifically mentioned ‘bloating’or similar (n = 9). Some mentioned cutaneous/subcutaneous sensations of rash/redness (n = 6),itching (n = 4), and tingling (n = 3). Most wereunable to carry out activities of daily living (ADL)during acute abdominal attacks (e.g. work, sociallife, household chores), perhaps needing to liedown or go to the emergency room (n= 19).

Oro-Facial-Pharyngeal-Laryngeal (OFPL) Locations

Participants who had experienced attacks inthe OFPL area (n = 22) distinguished betweenOFPL locations with difficulty, but spoke of theirconcern that facial or oral attacks might spreadto the throat with life-threatening consequences(i.e. asphyxiation). The worst symptoms were

breathing and swallowing difficulties (n = 5 each[internal] oro-pharyngeal-laryngeal location), dis-figurement (n = 9), edema (n = 7), and problemseating/drinking with swollen lips (n = 3, [external]facial location). Pain was generally described interms of ‘discomfort’ (US) or ‘bother’/‘nuisance’(Italy).

Spontaneous mentions: All 22 participantsmentioned edema. Ease of breathing was highlyrelevant to participants who had experiencedoro-pharyngeal-laryngeal attacks (n = 17/20 par-ticipants). Other aspects mentioned were as fol-lows: difficulties in swallowing (n= 13), difficulties/distortion of speech due to swollen lips or tongue(n= 9), or voice changes; and extreme embarrass-ment about facial disfigurement and associatedsocial stigma (n= 12). ADL were greatly compro-mised during a facial attack as participants wouldnot leave the house (n= 16), but unless a patientwent to the emergency room during a pharyngealattack, their ADL within the home were not af-fected greatly.

Peripheral Location

The most bothersome symptoms for partic-ipants who had experienced peripheral attacks

Table III. Key frequencies [n (%)] of spontaneous mentions of variables on each anatomical location visual analog scale instrument

VAS item Abdominal [n = 25] OFPL [n =22] Peripheral [n = 26] Urogenital [n = 17]

Pain 24 (96) 16 (72.7) 24 (92.3) 13 (76.5)

Swellinga 24 (96) 22 (100) 26 (100) 16 (94.1)

ADL 19 (76) 17 (77.3) 22 (84.6) 6 (35.3)

Illnessb 17 (68) 3 (13.6) NR 1 (5.9)

Bloatedb 9/15 (US only, 60) NR NR NR

Nausea 20 (80) NR NR 0

Hunger 13 (52) NR NR 1 (5.9)

Breathing NR 15 (68.2) NR NR

Speech NR 9 (40.9) NR NR

Swallowing NR 13 (59.1) NR NR

Using extremity NR NR 25 (96.2) NR

Urination NR NR NR 7 (41.2)

a There is no ‘swelling’ item on the Abdominal VAS instrument (US version), but 14 US participants mentioned the term ‘swelling’

spontaneously. There is little distinction between ‘swelling’ and ‘bloated’ in the Italian VAS instrument, the same root word, gonfiarsi, being

used in all four VAS instruments.

b Some US participants used wording very similar in meaning to both ‘ill’ and ‘bloated’ terms on the US Abdominal VAS instrument, e.g. gas

or distension.

ADL = activities of daily living; NR =not relevant (to a particular anatomical location); OFPL =oro-facial-pharyngeal-laryngeal; VAS = visualanalog scale.



(n= 26) were difficulty using extremity (n= 11), pain(n= 9), and edema (n= 8). Degree of pain varied ona continuum from ‘discomfort’ to ‘painful.’

Spontaneous mentions:All 26 participantsmen-tioned edema. Difficulty using the swollen limb(n = 25) affected ability to carry out ADL (n = 22);hindered ambulatory mobility occurred with footedema (n = 17); hindered dexterity (e.g. writing/typing, driving) occurred with hand edema; andparticipants expressed embarrassment aboutappearance/disfigurement (n = 5). Cutaneous/subcutaneous sensations were mentioned (n= 25).

Urogenital Location

For participants who had experienced uro-genital attacks (n = 17), the worst symptoms werepain/discomfort (n = 5), edema (n = 4), and diffi-culty urinating (n = 3). The experience of pain wasdescribed on the continuum from ‘painful’ to ‘notpainful at all.’

Spontaneous mentions: Sixteen participantsmentioned urogenital edema. Some referred topain or discomfort (n = 12); difficulties in urinat-ing (n = 7); difficulties with bowel movementswhen the edema spread (n = 2); difficulty wearingtheir usual clothes (n = 1); and cutaneous sensa-tions (n = 7). Participants mostly carried outADL as usual during a urogenital attack; discom-fort was associated with walking/sitting (n = 6)and cycling (n = 1).

Part 2: Cognitive Debriefing

Participants completed relevant location VASmeasures with reference to a typical acute attackfor which they would seek treatment. In general,VAS items were acceptable to respondents. Noparticipant had problems completing the VASlines, nor difficulty understanding the instructions,although six found the instructions unnecessarilyverbose. Table IV provides a brief summary ofthe cognitive debriefing results. Aspects commonto most VAS measures are reported below first

(items for ‘overall severity,’ ‘pain,’ ‘swelling,’ and‘illness’) followed by location-specific items.9

‘Overall Severity’ VAS (All Locations)

How severe are the angioedema symptoms nowfor this location? (No symptoms at all – Extremelydisabling): All participants considered this itemrelevant and generally comprehensible in context.Four Italians questioned the use of certain terms(e.g. ‘now’ and ‘location’ could be asking partic-ipants to insert time/day of questionnaire com-pletion), two Italians were initially uncertainabout themeaning. Response anchors were largelyacceptable to US participants, although ‘dis-abling’ was considered more appropriate to thePeripheral VAS instrument because peripheraledema was literally disabling. Six Italians com-mented that ‘damaged’ (= danneggiato) was un-suitable (the Peripheral VAS instrument is theonly Italian location VAS instrument on whichthe term appears) – potentially referring to cutsor wounds rather than effects of edema. Suchdifficulties may be due to translation issues.

‘Pain’ VAS (All Locations)

Do you have pain? (No pain – Extremely painful):All participants found this item relevant and easilyunderstood it. Most reported the term ‘pain’ ascovering the continuum of perceived discomfort.

‘Swelling’ VAS (All Locations except Abdominal)

Do you have any swelling? (No swelling – Ex-treme swelling): All participants found this itemrelevant and easily understood it. Although theUS Abdominal VAS instrument includes an itemabout ‘bloating’ rather than ‘swelling,’ there islittle distinction between ‘swelling’ and ‘bloating’in Italian.10

‘Illness’ VAS (All Locations except Peripheral)

How ill do you feel? (Not at all ill – Extremelyill): US participants had no difficulty under-standing the concept of ‘ill,’ however, there weredifferent interpretations depending on location: itwas seen as most relevant to abdominal attacks

9 Wording of both US and Italian VAS instruments is shown in the Appendix (SDC), but only the US wordingis shown here for ease of reference.

10 Both terms in Italian have the same root (gonfiarsi/gonfio).

120 McMillan et al.


(as a ‘bad’/‘sick’ feeling), and least relevant forurogenital attacks. Some (n = 3) considered ‘ill’ asan infection or cancer and irrelevant to HAE, butstill understood the item as indicating attack se-verity. Italian participants generally had no nota-ble comments, other than that questions on feelingill (question 1) and experiencing pain (question 2)were asking essentially the same question, both

referring to pain11 [abdominal and OFPL loca-tions (n= 1), urogenital (n= 4)].

Abdominal-Specific VAS Items (‘Bloated,’ ‘Nausea,’‘Hunger,’ ‘Desire to Eat’)

The majority of participants found theabdominal-specific items relevant and easily un-derstood them.

Table IV. Summary of visual analog scale cognitive debriefing results

Relevant (n) Comprehensible (n) Comments

Title and instructions (n =27) NA 27 Instructions comprehensible but rather verbose (n= 6). Title (HAEattack location) generally comprehensible (n =17 US participants

only)

VAS items common to all four

attack locations (n= 27)

Overall severity 27 17/17 (US)

7/10 (Italy)

‘Now’, ‘location,’ and ‘severe’ (in the context of ‘disabling’) were

queried by Italian participants

Danneggiato (Italian Peripheral VAS instrument only) suggested

as inappropriate (but comprehensible) term (6/10 Italians)

Illness 25 27 Different interpretations of ‘ill’ depending on attack location,

e.g. nausea. ‘Ill’ not relevant if term considered as indicating an

infection or serious condition such as cancer

Pain 27 17/17 (US)

7/10 (Italy)

‘Pain’ covers continuum of perceived discomfort

‘Ill’ and ‘pain’ can be similar concepts in Italian (n= 3)

Swellinga 27 27 None

Abdominal-specific items (n=25)

Bloated (US only) 12/15 15/15 ‘Full’/‘empty’ more appropriate to food (n= 2, US)

Nausea 25 25 Relevant but perhaps redundant as covered by ‘illness’ (n =2)

Hunger 10/15 (US)

10/10 (Italy)

25/25 Not relevant because patients avoid eating during HAE attacks

(n= 5)

Desire to eat (Italy only) 8/10 10/10 Superfluous in addition to ‘hunger’ (n= 2)

OFPL-specific items (n= 22)

Breathing 22 22 More relevant to pharyngeal/laryngeal attacks than facial (n =2)

Speech 22 22 None

Swallowing 22 22 More relevant to pharyngeal/laryngeal attacks than facial (n =1)

Peripheral-specific items (n= 26)

Using extremity 26 26 Most participants used the term ‘hands and feet’ rather than

‘extremities’

Urogenital-specific items (n= 17)

Hunger 6 17 Appetite not affected vs avoiding eating/drinking to minimize

need to urinate/defecate

Nausea 8 17 None

Urination 17 17 None

a The term ‘bloated’ was used in the US Abdominal VAS instrument.

HAE =hereditary angioedema; NA = not applicable; OFPL = oro-facial-pharyngeal-laryngeal; VAS = visual analog scale.

11 This might be because the Italian male (as in quanto male sente? =how ill do you feel?) has multiple inter-pretations in Italian: ill, evil, pain, or harm.



Main comments: For ‘bloated’ (US VAS in-strument only), response anchors ‘full’/’empty’were considered more appropriate to food (n = 2).‘Nausea’ was redundant as it was covered by ‘ill-ness’ (n= 2). ‘Hunger’ was not very relevant becausepatients avoided eating during abdominal attacksfor fear of vomiting (n = 8). For ‘desire to eat’(Italian VAS instrument only) or ‘hunger,’ one orother item was superfluous on the same Italianinstrument (n = 2).

Missing symptoms: Results were as follows: nomissing symptoms (15/25 participants); light-headedness [feeling faint/dizzy/passing out] (n= 5);vomiting (n= 2); and mobility problems, diarrhea,or dehydration (n = 1 each). However, vomiting(action) is arguably latent content for ‘nausea’(feeling), a more extreme symptom on the con-tinuum, and dehydration and light-headedness di-rectly result from vomiting. No Italian mentionedfatigue when asked about missing abdominalsymptoms, despite six of ten Italians indicating(during Part 1) that fatigue was a feature ofabdominal attacks.

OFPL-Specific VAS Items (‘Breathing,’ ‘Speech,’‘Swallowing’)

There were no problems with comprehension;all participants considered questions relevantunless otherwise stated.

Main comments: Two items were consideredrelevant for pharyngeal/laryngeal attacks but notfor facial attacks: ‘breathing’ (n = 2) and ‘swal-lowing’ (n = 1).

Missing symptoms: Results were as follows: nomissing symptoms (16/22 participants), but visualproblems and facial disfigurement (n = 2 each) –both direct effects of edema and not symptomsper se.

Peripheral-Specific VAS Items (‘Using Extremity’)

All participants understood and found theconcept relevant, but most used the term ‘handsand feet’ rather than ‘extremities’ or ‘peripheral,’because acute attacks began, or weremost severe, inthe outer extremities rather than the arms and legs.

Missing symptoms: Results were as follows:none (22/26 participants); skin irritation (n = 2);and changes in skin’s appearance, and feelings of

limb heaviness (n = 1 each). These symptoms canbe viewed as latent content for ‘swelling’ or ‘usingextremity.’

Urogenital-Specific VAS Items (‘Nausea,’ ‘Urination,’‘Hunger’)

There were no problems with comprehensionbut ‘nausea’ (n= 9) and ‘hunger’ (n= 11) were con-sidered irrelevant to participants’ urogenital attacks.There were different interpretations of hunger inthis context (i.e. whether appetite was affected byattacks or drinking/eating were avoided to mini-mize the need to urinate or defecate).

Missing symptoms: Results were as follows:none (11/17 participants); mobility problems (n= 2);and bowel movements, sexual difficulties, irrita-tion, or pain on urination (n = 1 each). All of thesecan be viewed as latent content for ‘swelling.’

Comments on Semantics and Direction ofResponse Options

There were some comments from Italian re-spondents regarding grammatical points. Theitems ‘pain,’ ‘nausea,’ ‘hunger,’ and ‘swelling’ (e.g.‘Do you have any swelling?’) require a dichoto-mous (yes/no) response rather than a scale; ‘notdamaged’ is not the opposite of ‘extremely dis-abling’ (Peripheral VAS instrument only). OneItalian noticed the reversed response options of‘urination,’ ‘breathing,’ and ‘using extremity,’the only item on their respective instruments(Urogenital, OFPL, and Peripheral) with severesymptoms at the lower (0mm) end of the scale.

The Title and Instructions

Only US participants were asked specificallyabout the title of the VAS instruments, and themajority had not noticed the title until it waspointed out to them.� Abdominal (n = 15): the medical term ‘abdom-

inal’ was well understood by all participants,though most would use ‘stomach.’

� Peripheral (n = 16): some thought that ‘pe-ripheral’ might refer to vision (n = 6), and allwould prefer the term ‘extremities’ or specifi-cally ‘hands’/‘feet’. ‘Hands and feet’ were pre-ferred to ‘arms and legs’ because acute attacksbegan, or were most severe, in the outerextremities.

122 McMillan et al.


� Oro-facial-pharyngeal and/or laryngeal (n= 13):although most participants did not know theexact meaning of the technical terms, they allunderstood the title correctly to refer to facialand throat areas.

� Urogenital (n = 9): the title was not associatedwith any comprehension problems but otherterms suggested were ‘privates’ and ‘genital area.’All US and Italian participants found the in-

structions comprehensible, but six commentedthat the instructions were rather verbose.

Repeat Administration of VAS instruments in ClinicalTrial Situations

Most participants (n = 16) reported that theVAS instruments were suitable for frequent ad-ministration during treatment for an acute HAEattack in a clinical trial, being short and easilyunderstood. Completion may be problematic ifthe writing hand was swollen, but the other handcould be used easily to mark the VAS line. Severepain might be distracting, but 11 participants re-ported that it would not be a problem for them.

Discussion

The aim of this interview study was to assessthe content validity and suitability of four con-dition-specific VAS instruments for use in clinicalstudies of rhC1INH treatment. Semi-structuredexploratory interviews, conducted in the US andItaly, investigated the experience of acute angio-edema attacks in 27 adults with HAE. Cognitivedebriefing of the VAS instruments was under-taken with reference to a typical acute attack forwhich participants would seek treatment. To ourknowledge, this study is the first to report qual-itative data about the patient’s experience ofHAE.

A representative sample of 27 adults withHAE was recruited (a highly satisfactory numbergiven the rarity of HAE) and percentages ofparticipants experiencing anatomical attack lo-cations were similar to those found in a Germansurvey.[21] Few new symptoms/comments on theVASs emerged after the first 12 US interviews,indicating good saturation. Little difference wasfound between US and Italian participants’ ex-perience of HAE.

Symptoms mentioned by participants in theexploratory interviews were similar to those(where reported) in the above-mentioned Germansurvey, e.g. voice changes (laryngeal edema),pain, or difficulties at micturition (urogenitaledema).[21] As expected, the present study foundedema and/or pain were the predominant HAEattack symptoms in all anatomical locations,[1,3]

supporting their inclusion in the VAS instru-ments. Participants’ spontaneous mentions of‘swelling’ ranged from 100% (OFPL and peri-pheral locations) to 94% (urogenital). Edema wasthe most bothersome symptom in facial attacks,with many mentioning concerns (embarrassmentand perceived stigma) over facial appearance(disfigurement). Swollen lips caused problemswith speech, eating, and drinking and respondentsfeared that edema might spread to the throat andbecome life threatening. In Part 2, all participantsfound the ‘swelling’ item highly relevant and easilyunderstood it. Although there is no specific ‘swel-ling’ VAS on the US Abdominal measure, the‘bloated’ VAS was found to be a fair alternative.

Pain or discomfort were mentioned sponta-neously by 96% (abdominal location) to 73%(OFPL) of participants, with descriptions of paincovering a continuum up to moderate or severe,supporting findings of other studies.[5] In Part 2,all participants found the ‘pain’ VAS item highlyrelevant and easily understood it. However, theterm ‘ill’ was rarely used spontaneously – it tendedto be associated with an infection/nausea/vomiting.Hence, it is most appropriate to abdominal attacks.

The location-specific items ‘nausea’ and ‘hun-ger’ (Abdominal VAS instrument), ‘breathing,’‘speech,’ ‘swallowing’ (OFPL), ‘using extremity’(Peripheral), and ‘urination’ (Urogenital) were allconsidered relevant and comprehensible, (spon-taneous mentions ranged from 96% [‘using ex-tremity’] to 41% [‘speech’ and ‘urination’]). Fewparticipants would use the term ‘extremity,’ andsome questioned ‘hunger’ as they do not eatduring an abdominal attack. The term ‘damaged’(= danneggiato) was considered unsuitable (astranslated on the Italian Peripheral VAS instru-ment). There was little support for includingeither ‘hunger’ or ‘nausea’ on the UrogenitalVAS instrument, or ‘desire to eat’ (additional to



‘hunger’) on the Italian Abdominal VAS instru-ment only.

Any additional symptoms mentioned wereoften considered less important by respondentsin a brief measure, and are largely coveredby existing VAS items, e.g. vomiting falls under‘nausea’ (Abdominal). Thus, all serious and signi-ficant symptoms mentioned by most participantswere covered in the VAS instruments, directly orindirectly.

ADL were often severely compromised byabdominal, OFPL, and peripheral attacks, andmost participants stayed at home during facial at-tacks due to embarrassment about their appear-ance. Urogenital attacks, however, had little effecton ADL. Thus, HAE attacks could negativelyimpact on ability to work or on education, as wasfound in a recent large survey of HAE patients.[5]

In general, participants’ descriptions of symptomsand their effects on ADL supplied much richerdata12 than are currently available in the literature.

The ‘overall severity’ VAS is intended to pro-vide the patient’s perspective of angioedemaattack severity and its effect on his/her life, con-sidering all relevant symptoms for that location.Although the term ‘disabling’ was rarely usedspontaneously, participants instead frequentlyreferred to their ADL, which could be viewed aslatent content for ‘disabling.’ Respondents gen-erally did not report difficulties completing the‘overall severity’ item and understood the item asa summation of symptoms. The ‘overall severity’item could stand alone as a single item measure,especially if there were particularly severe symp-toms at VAS completion. Although several Italianparticipants remarked on the unsuitability of‘damaged’ (= danneggiato) on the Peripheralmeasure, this did not preclude item understanding.

Titles of the HAE VAS measures used medicalterminology for the attack locations. Only USparticipants were asked specifically about the titleof the VAS instruments in this study, and themajority did not comment on the title until it was

pointed out to them. In general, the great ma-jority of participants understood the terminologyalthough they might use more colloquial termswhen referring to their attack locations. TheseVASmeasures, when originally constructed, wereintended for use in clinical trials when patientswere experiencing an acute HAE attack. Duringsuch clinical trials, the investigator (not the pa-tient) confirmed pre-treatment that a locationwas eligible for assessment and scored the se-verity of angioedema attack symptoms. At thesame time, the patient experiencing the attack(s)was given the VAS instrument appropriate totheir attack location(s), on which they indicatedthe severity of angioedema at the attack location.Thus, in practice, patients undergoing an HAEattack would be in no doubt about the location ofthe attack, would be given the VAS instrumentspecific to that location, and would not necessar-ily need to understand the VAS instrument title.Medical personnel were always on hand to an-swer any questions.

This study provided evidence for good contentvalidity for the items on these HAE VAS mea-sures, with a strong relationship between numbersof spontaneous mentions and almost all VASitems, except ‘nausea’ and ‘hunger’ (UrogenitalVAS instrument), and ‘desire to eat’ (ItalianAbdominal VAS instrument only). The suit-ability of the VASs in both the US and Italy issome indication of their cross-cultural validity.Asking participants to consider a hypotheticalscenario (i.e. experiencing an imaginary acuteangioedema attack) when completing the VASsis generally considered problematic[22] but thispragmatic approach was found to be acceptableto participants, all of whom had vivid recol-lections of previous acute attacks.13 Participantsreported the VAS instruments as simple andquick to complete, suitable for frequent repeatadministration during acute HAE attacks, thus,the VAS instruments are suitable for use in clin-ical trials. However, they may be improved. The

12 See figure S1 in the SDC.

13 Given the frequency and severity of HAE attacks, recall was not expected to be problematic, as diabetesmellitus studies have noted that recall of severe hypoglycemic events (of similar magnitude and distress to HAEattacks) is robust up to 1 year or more.[23]

124 McMillan et al.


VAS instructions might be simplified. The OFPLinstrument could be split into two measures (fa-cial and oro-pharyngeal-laryngeal) so that allsymptoms are pertinent to each location (but thiswould increase respondent burden). Minor issueson the Italian instruments could be improved toavoid the possibility of misinterpretation.

Initial work on the psychometric properties ofthe VASs has been completed (unpublished).These findings demonstrate that the reliability andvalidity of the VAS instruments are consistentwith the properties of VAS measures observed inacute pain studies[24,25] as well as in other studiesrecently conducted in HAE patients.[26] The VASinstruments could be useful in the clinical setting,subject to further research (for example concern-ing the timing and frequency of their use).

Conclusions

This interview study explored patients’ expe-rience of HAE attacks, offering the first pub-lished qualitative descriptions of symptoms andtheir (often) debilitating impact on ADL. It hasdemonstrated comprehensive content and goodcontent validity of the four anatomical locationHAE VAS instruments with comprehensibility ofitems/responses. The HAE VAS instruments aresuitable for use in assessment of acute angio-edema symptoms in rhC1INH clinical studies.

Acknowledgments

This study and manuscript preparation were funded by Phar-ming Technologies BV who were involved in the design and con-duct of the study, and the review and approval of the manuscript.

CVM is a Consultant Health Psychologist at AHP Re-search and JS is Director of Research at AHP Research, anindependent research company that received funding fromPharming to conduct this study and received consultancy feesrelated to contributions to other aspects of Pharming’s clinicalstudies. AR is a Medical Director at Pharming TechnologiesBV and has also received stock options in the company. GH isan independent consultant who was engaged by Pharming andreceived fees and expenses from the study sponsor throughoutthe conduct and reporting of the study. MC has or has hadconsultancy agreements with Pharming, Jerini, andDyax and ison the advisory board for Jerini/Shire, CSLBehring, Pharming,Viro Pharma, and Dyax. Universita degli Studi di Milano(= TheUniversity ofMilan) received a payment fromPharmingfor performing the study. LB has a consultancy agreement withPharming for running research and development activities.

The authors acknowledge valuable assistance from staff atall clinics (Family Allergy & Asthma Center P.C., Atlanta,GA, USA; Asthma, Allergy& Immunology Clinical ResearchUnit, Tampa, FL, USA; Allergy, Asthma & ImmunologyClinic P.A., Irving, TX, USA; and Dipartimento di ScienzeCliniche ‘‘Luigi Sacco,’’ Ospedale Luigi Sacco, Milan, Italy)where people with HAE were recruited and interviews carriedout. The authors thank Adele Succetti for conducting the in-terviews in Italy and all 27 participants for sharing their ex-periences and insights intoHAE and contributing their time toaid our understanding of this condition.

GH and AR conceived the need for the interview study,which was designed by JS, CVM, AR, and GH and coordi-nated by CVM.MC reviewed the study protocol. JS and CVMconducted the interviews in the US and LB coordinated theinterviews in Italy. CVM and JS analyzed the data and wrotethe report upon which this manuscript is based. CVM draftedthe manuscript, JS and AR contributed to the manuscriptdevelopment, and all authors read and approved the finalmanuscript. CVM is the guarantor of the overall content ofthis article.

References1. Agostoni A, ygoren-Pursun E, Binkley KE, et al. Hereditary

and acquired angioedema: problems and progress – pro-ceedings of the Third C1 Esterase Inhibitor DeficiencyWorkshop and beyond. J Allergy Clin Immunol 2004 Sep;114 Suppl. 3: S51-131

2. Davis III AE. The pathogenesis of hereditary angioedema.Transfus Apher Sci 2003 Dec; 29 (3): 195-203

3. Zuraw BL. Clinical practice: hereditary angioedema. N EnglJ Med 2008 Sep 4; 359 (10): 1027-36

4. Gompels MM, Lock RJ, Abinun M, et al. C1 inhibitor de-ficiency: consensus document. Clin Exp Immunol 2005Mar; 139 (3): 379-94

5. LumryWR, CastaldoAJ, VernonMK, et al. The humanisticburden of hereditary angioedema: impact on health-relatedquality of life, productivity, and depression. Allergy Asth-ma Proc 2010 Sep; 31 (5): 407-14

6. Myles PS, Urquhart N. The linearity of the visual analoguescale in patients with severe acute pain. Anaesth IntensiveCare 2005 Feb; 33 (1): 54-8

7. Stadler M, Schlander M, Braeckman M, et al. A cost-utilityand cost-effectiveness analysis of an acute pain service.J Clin Anesth 2004 May; 16 (3): 159-67

8. Gerlinger C, Schumacher U, Faustmann T, et al. Defining aminimal clinically important difference for endometriosis-associated pelvic pain measured on a visual analog scale:analyses of two placebo-controlled, randomized trials.Health Qual Life Outcomes 2010 Nov 24; 8 (1): 138

9. Kelly AM. The minimum clinically significant difference invisual analogue scale pain score does not differ with severityof pain. Emerg Med J 2001 May; 18 (3): 205-7

10. Cicardi M, Banerji A, Bracho F, et al. Icatibant, a newbradykinin-receptor antagonist, in hereditary angioedema.N Engl J Med 2010 Aug 5; 363 (6): 532-41

11. U.S. Department of Health and Human Services FDACenter for Drug Evaluation and Research, U.S. Depart-ment of Health and Human Services FDA Center for



Biologics Evaluation and Research, U.S. Department ofHealth and Human Services FDA Center for Devices andRadiological Health. Guidance for industry. Patient-reported outcome measures – use in medical product de-velopment to support labeling claims: draft guidance.Health Qual Life Outcomes 2006; 4: 79

12. Zuraw B, Cicardi M, Levy RJ, et al. Recombinant humanC1-inhibitor for the treatment of acute angioedema attacksin patients with hereditary angioedema. J Allergy ClinImmunol 2010 Oct; 126 (4): 821-7

13. Vernon MK, Rentz AM, Wyrwich KW, et al. Psychometricvalidation of two patient-reported outcome measures to as-sess symptom severity and changes in symptoms in heredi-tary angioedema. Qual Life Res 2009 Sep; 18 (7): 929-39

14. Guyatt GH, Feeny DH, Patrick DL. Measuring health-related quality of life. Ann Intern Med 1993; 118: 622-9

15. Bowling A.Measuring health: a review of quality of life mea-surement scales. 3rd ed. New York (NY): Open UniversityPress, 2005

16. Willis G. Cognitive interviewing: a tool for improvingquestionnaire design. Thousand Oaks (CA): Sage Pub-lications, 2005

17. Rothman M, Burke L, Erickson P, et al. Use of exist-ing patient-reported outcome (PRO) instruments andtheir modification: the ISPOR Good Research Practicesfor Evaluating and Documenting Content Validity forthe Use of Existing Instruments and Their ModificationPRO Task Force Report. Value in Health 2010; 12 (8):1075-83

18. Joffe H, Yardley L. Content and thematic analysis. In:Marks DF, Yardley L, editors. Research methods forclinical and health psychology. London: Sage, 2004: 56-68

19. Craig TJ, Levy RJ, Wasserman RL, et al. Efficacy of humanC1 esterase inhibitor concentrate compared with placebo in

acute hereditary angioedema attacks. J Allergy Clin Im-munol 2009 Oct; 124 (4): 801-8

20. Bouillet L, Longhurst H, Boccon-Gibod I, et al. Diseaseexpression in women with hereditary angioedema. Am JObstet Gynecol 2008 Nov; 199 (5): 484

21. Bork K,Meng G, Staubach P, et al. Hereditary angioedema:new findings concerning symptoms, affected organs, andcourse. Am J Med 2006 Mar; 119 (3): 267-74

22. Speight J, Reaney M, Barnard KD. The use of hypotheticalscenarios and importance weightings when measuring theimpact of diabetes on quality of life: a response to Brose etal. Diabet Med 2010; 26 (10): 1077-9

23. Pramming S, Thorsteinsson B, Bendtson I, et al. Symptom-atic hypoglycaemia in 411 type 1 diabetic patients. DiabetMed 1991 Apr; 8 (3): 217-22

24. Gallagher EJ, Bijur PE, Latimer C, et al. Reliability andvalidity of a visual analog scale for acute abdominal pain inthe ED. Am J Emerg Med 2002 Jul; 20 (4): 287-90

25. Todd KH, Funk KG, Funk JP, et al. Clinical significance ofreported changes in pain severity. Ann Emerg Med 1996Apr; 27 (4): 485-9

26. Food and Drug Administration. Firazyr (icatibant) in-jection for the treatment of acute attacks of hereditaryangioedema (HAE) in patients 18 years of age and older:briefing document [report no. NDA 022-150]. 2011 [on-line]. Available from URL: http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/Pulmonary-AllergyDrugsAdvisoryCommittee/UCM260020.pdf [Accessed 2011 Jul 5]

Correspondence: Dr Carolyn McMillan, AHP Research,16 Walden Way, Hornchurch RM11 2LB, UK.E-mail: [email protected]

126 McMillan et al.


Documents

Content Validity of Visual Analog Scales to Assess Symptom Severity of Acute Angioedema Attacks in Adults with Hereditary Angioedema