A modified Delphi panel ‒ a qualitative research tool often used in healthcare to obtain consensus on a specific topic among a panel of experts,16 was conducted considering recommendations of Trevelyan 201517 and Nasa 2021,18 in order to assess assumptions and obtain consensus in areas where information is limited or lacking.19

The study protocol was developed a priori and the Research Ethics Committee of the Federal University of Goiás in Goiania, Brazil, granted ethical approval for this investigation in December 2022 (#5.822.872). Electronically signed consent was obtained from those who agreed to participate, and the participants were reimbursed for their time in the amount of R$500.00.

Members identified by referral from medical societies, specialists in the field, and research groups were invited to be on the Delphi panel. Accordingly, a diverse group of specialists with experience in HZ care from the public and/or the private healthcare system, in the various levels of care, and from all 5 macro-regions of the country, were invited to participate in the expert panel. There are no standard recommendation with respect to the number of panel/experts and it varies from 10 to 1000. However, most of the studies are conducted with 20 to 50 panelists, and the number may be even smaller when the group is highly homogeneous.15,18 The panel was composed of 20 members representing infectious disease specialists (n = 8), dermatologists (n = 5), neurologists (n = 2), ophthalmologists (n = 2), geriatricians/clinicians (n = 2), and physiatrists (n = 1). Panel members were informed about the objective and procedure of the study and those who agreed to participate were asked to sign an informed consent.

The Delphi questionnaire was structured by areas of expertise and considering a variety of relevant clinical presentations including a typical HZ case, PHN case, ophthalmic zoster, and neurologic zoster in immunocompetent individuals (Fig. 1). This flowchart was presented early on the questionnaire to all panel members, and depicted clinical presentations of HZ to be considered when responding to the questionnaire, along with the diagnostic therapeutic approaches recommended by current international guidelines.10–12,20–25

Fig. 1.

Flowchart Depicting Clinical Presentations of HZ considered in the Delphi Panel, by guideline recommended diagnostic and therapeutic approaches. Brazil, 2023.

(0.59MB).

Parameters and questions included in the questionnaire were defined considering both a literature review conducted in April 2023, and available data from the National health information systems (public and private) in Brazil by the project team. The literature search updated a previous review conducted by Bardach et al. 2021,5 replicating terms used and search strategy. Data elements which was not available, limited or inconsistent were considered and included in the Delphi panel assessment.

These included estimates of proportion of HZ patients with PHN or recurrent PHN; duration of HZ and PHN episode; proportion of patients with selected complications (dermatologic, ophthalmic, neurologic, or other complication); proportion of patients requiring referral to other specialties; proportion of patients requiring hospitalization; and proportion of patients reporting work absenteeism, by HZ clinical presentations considered. Absenteeism was defined as the proportion of patients who were away from occupational activities due to prescribed medical leave. Average number of lost days of work was estimated considering the duration of prescribed medical leave.

In addition, diagnostic and therapeutic management of patients was also assessed, diagnostic modalities (clinical only, clinical and laboratory) and types of laboratory diagnostic tests used; average time of medical follow up; laboratory exams and medical visits during patient follow up; type of antivirals and number of days used; therapy for pain management and number of days used, and other medications, also by HZ clinical presentations considered.

Questions were stratified by public vs. private healthcare system and age group (50‒59, 60‒69, 70‒79 and 80+ years of age).

Questionnaire was pilot tested with the project team and a guest medical expert. The online questionnaire was structured and made available via SurveyMonkey®. Questionnaires were comprised of 140 and 69 questions, in the 1st and 2nd Delphi panel rounds, respectively.

All participants who agreed and signed the informed consent were sent the online questionnaire and were given 20 days to fill out and reply to their responses. A reminder was sent to panel members 4 days before the deadline. Responses was anonymous. All panel members were asked to respond to the questionnaire according to their expertise. Infectious diseases specialists, dermatologists, and geriatricians/clinicians provided answers regarding the typical case of HZ and PHN. They also provided answers for the neurological and ophthalmic cases if they had experience with these. Similarly, neurologists provided answers related to PHN and the neurological case, while ophthalmologists responded only to the ophthalmic case.

To answer the questions, experts were asked to consider a population of 100 patients with HZ and indicate the average proportion of patients affected by the simulated scenarios or the average number of resources used in each described situation. Panel members were also provided with the opportunity to comment and propose new items if needed.

As it was required that a consensus be reached, two rounds of the Delphi panel took place. After responding to the questionnaire, panel members received a summary of the group replies based on which individual experts modified their responses. This process was repeated until an expert consensus was reached. Consensus was defined as a percentage of agreement based on a predefined cut-off value usually selected arbitrarily18 – which was defined as 75 % or more for the 1st round. If there was no consensus in the 1st round, another round was held and a new questionnaire considering only questions for parameters to which no consensus was reached was sent to panel members.

Consensus threshold was defined as 70 % or more in the 2nd round. Questions were presented with reported average values from the literature (when available) or the average and minimum/maximum values obtained from the 1st round of the panel. Response options were presented with a 4-point Likert-type scale (1 = strongly disagree, 2 = partial disagree, 3 = partial agree, and 4 = strongly agree). The experts had the chance to reflect on their previous responses in light of these scores and to revise them when they disagreed with the midpoint. If consensus was not reached in the 2nd round, another round could be held to explore the group’s divergence points.

Descriptive analysis was conducted, and summary statistical measures are reported as average and minimum and maximum values for numerical variables, and percentages for proportions. Results were stratified by age-groups and healthcare system (public and private). For consensus assessment, inter-respondent agreement was calculated considering intra class correlation coefficient. If 1–2 panelists differed in their answers in relation to the other panelists by > 50 % with extreme values, their experience was assessed considering number of HZ patients they had treated per year and their answers were excluded if they had no significant experience with that specific issue.

Among the 20 experts who agreed to participate in the 1st round of the Delphi panel, 19 completed the questionnaire. For the 2nd round, only the respondents of the 1st round were considered, and 17 experts completed the questionnaire. The detailed steps, procedures and timeline of the Delphi panel are presented in Supplement Table 1, whereas the characteristics of Delphi Panel members, by geographic location, level of care and type of healthcare system where they provide care are presented in Table 1. Over 50 % of the panel members were from the Southeast region of Brazil. Panel members reported a range of 3 to 50 HZ patients treated in the previous year. The average time for questionnaire completion was 60 and 28 minutes in the 1st and 2nd rounds, respectively.

Fig. 1 depicts clinical presentations of HZ considered in the Delphi Panel. Parameters of disease progression, complications and work absenteeism are presented in Table 2. The proportion of patients with episodes of PHN and PHN recurrence increased with age, from 4 % and 3 %, respectively, for the 50–59 age group, to 14 % for the 80 and over age group. Similar observations were noted for patients presenting with dermatological and ophthalmic complications.

Across age groups, the proportion of patients absent from work due to HZ and PHN was 50 % and 30 %, respectively. Among these, the average number of workdays lost were 29 and 39 due to HZ and PHN, respectively. These estimates increased in patients with complicated zoster, reaching 68 % in patients with ophthalmic complications, and 48 % with neurological complications. The number of additional days of work lost due to these complications was 18 and 39, respectively. Absenteeism estimates did not differ by age groups.

Table 3 shows the number of consultations in outpatient care and subsequent proportion hospital admissions for a standard case of HZ and PHN in the private healthcare system as determined by the Delphi panel. The number of medical consultations for HZ and PHN varied from 2 to 3 and 4 to 6 increasing with age group, respectively, and did not differ significantly by healthcare system. The proportion of patients referred to another medical specialty also increased with age and was higher in private when compared to patients in the public healthcare system. The proportion of patients requiring hospitalizations increased with age from 1 % for the 50–59 age group to 8 % for the 80 and over age group in both health systems.

Regarding HZ diagnosis and clinical management, main types of resources used in the private and public healthcare systems are presented in Table 4. HZ diagnosis was predominantly confirmed by clinical criteria in more than 90 % of HZ typical cases in both public and private healthcare systems. Regarding antiviral use, typical HZ patients are predominantly treated with acyclovir (95 %) public health system, while in the private sector valacyclovir is mostly used (80 %).

Pain management was initially addressed with dipyrone and codeine for around 60 % of typical HZ cases, while prednisone and tramadol were used in 30 % of cases. For PHN cases, the use of amitriptyline was reported in 56 % of patients in the public healthcare system and 40 % of patients in the private system, followed by gabapentin (Neurontin) in around 45 % of patients in both systems. The use of codeine and tramadol was also mentioned in PHN for up to 42 % and 36 % of cases, respectively. Pregabalin was indicated in 27 % and 58 % of patients from the public and private healthcare systems respectively (Table 4).