Development, validation and feasibility of a Patient Satisfaction Questionnaire for evaluating the quality performance of a diagnostic small fibre neuropathy service: A qualitative study

Abstract Introduction and Aim Small fibre neuropathy (SFN) is a peripheral neuropathy, leading to neuropathic pain and autonomic dysfunction. An evidence‐based standardized patient diagnostic SFN service has been implemented in the Netherlands for improving patient‐centred SFN care. However, the quality of care of this diagnostic SFN service has never been assessed from a patient perspective. The aim of this study was to develop and validate an SFN‐Patient Satisfaction Questionnaire (SFN‐PSQ) to measure the quality performance of a standardized diagnostic SFN service. Methods A descriptive qualitative study to create the SFN‐PSQ was performed using the (COREQ (Consolidated Criteria for Reporting Qualitative Research) checklist. For item generation and content development, domains and/or items from validated PSQs were selected. The content development and content validity were performed using a Delphi method with SFN expert caregivers with different backgrounds. By using the three‐step‐test method in individual cognitive interviews, the content validity by patients was finalized. Results In one online Delphi panel round, the content of the first concept of the SFN‐PSQ was validated, which resulted in the second concept of the SFN‐PSQ. From July 2019 till March 2020, nine patients consented to participate in the individual cognitive interviews. The most significant changes of the new questionnaire were adding domains and items concerning the waiting list, the diagnostic services and consultation by the hospital psychiatrist. Also, a differentiation was made for both an inpatient and outpatient diagnostic SFN service. Furthermore, the clarity and intelligibility of the domains/items were improved, resulting in an increased comprehension of the SFN‐PSQ. Ultimately, the new developed SFN‐PSQ consisted of 10 domains and 51 items, suitable for measuring patient satisfaction of the neurological analysis in patients with SFN. Conclusion Through item generation, expert opinions and interviews with patients, the SFN‐PSQ was developed and validated, and feasibility was confirmed. The structure of the questionnaire, based on the logistic and diagnostic SFN pathway, could be used as a model in other hospitals to improve the quality, continuity and access of SFN care and other chronic diseases taking into account potential cross‐cultural differences. Patient or Public Contribution Caregivers were involved in the item generation and content development of the questionnaire. Patients were directly involved in testing the content validity and feasibility of the SFN‐PSQ. Clinical Trial Registration Not applicable.


| INTRODUCTION
As a peripheral sensory neuropathy, small fibre neuropathy (SFN) is a neuromuscular disorder. 1SFN affects the thinly myelinated and unmyelinated nerve fibres, leading to severe neuropathic pain and autonomic dysfunction. 1,2SFN interferes with daily and physical functioning, and thereby contributes to a reduced quality of life. 3The prevalence rate of SFN is 53/100,000 inhabitants, 4 but due to underdiagnosing SFN as a result of difficulties in recognizing SFN symptoms, its incidence and prevalence may be underestimated. 5rrently, Dutch hospitals have not implemented a standardized and integrated diagnostic SFN service.Not all diagnostic costs (i.e., skin biopsy) are reimbursed by Dutch health insurance, resulting in use of expensive infrastructure, an increase of patients' burden and substantial health-related costs.Implementation of a standardized, integrated SFN patient journey is necessary for potentially costsaving and efficient, accurate diagnosis of SFN.In addition, it provides a basis for etiological work-up and treatment decisions.
SFN interferes with daily and physical functioning, and thereby contributes to a reduced quality of life. 37][8] Recently, two studies examined the accessibility and patient satisfaction of occupational therapy services for patients with hereditary transthyretin amyloidosis, 9 and teleneurology clinics for patients with polyneuropathy, 10 but no earlier research on patient satisfaction of the interinstitute validated and well-defined diagnostic SFN approach has been performed.Yearly, 500 new patients are neurologically analysed at the diagnostic SFN service, and the majority visit the NDCU.Only patients who are living near the hospital visit the OC.
When test results are available, patients are discussed in the multidisciplinary team with all potential stakeholders (neurologists, pain specialists, psychiatrists, psychologists, geneticists, nurse practitioner [NPs], nurses, physiatrists) for patient-centred advices and treatment options based on evidence-based practice.As part of the diagnostic SFN service, anxiety and depression are assessed using the Hospital Anxiety and Depression Scale (HADS) questionnaire, which consists of two subscales: one measuring anxiety (HADS-A) and the other measuring depression (HADS-D). 11In case of a HADS score of ≥14, a patient consultation by the psychiatrist is planned.
3][14][15] Themes of patient satisfaction identified in the healthcare context are provider attitude, technical competence, accessibility and efficacy. 168][19][20] There are various instruments for measuring patient satisfaction in patients with chronic pain, [21][22][23] but these instruments often measure the quality of medical care and pain treatment goals, instead of patient satisfaction on the diagnostic process.One validated and feasible core questionnaire for the assessment of patient satisfaction (Core Questionnaire for the assessment of Patient Satisfaction Questionnaire [COPS]) in academic hospitals in the Netherlands has been developed, 24 with the aim of comparing benchmarking patient satisfaction results among the hospitals.Furthermore, since 2009, the Dutch Federation of University Medical Centers (NFU) has been using the Consumer Quality Index Questionnaire Outpatient Hospital Care (CQI-QOHC) for measuring customer experiences in hospital admissions. 25However, all these Patient Satisfaction Questionnaires (PSQs) are designed to measure the overall quality of general hospital and medical care, but do not cover domains of specific care pathways components, including diagnostic services.
Hence, to improve the quality of SFN care by measuring the patient experience and satisfaction related to the diagnostic SFN service, development of a questionnaire encompassing all relevant aspects of the service is needed.The aim of this study was to develop and validate an SFN-PSQ for evaluating the quality performance of the diagnostic SFN service from the patients' perspective.For creating the content of the SFN-PSQ, the Dutch guideline for qualitative methods was used. 26As suggested in this tool, the expert opinion (i.e., using a Delphi Method), literature, existing PSQs and other additional sources were used to create items that reflect the concept of patient satisfaction of the diagnostic SFN service.Furthermore, individual cognitive interviews were performed for the content validation of the SFN-PSQ. 27The subsequent steps of the content development of the SFN-PSQ are explained in detail in the text below.

| Phase 1a: Item generation and content development of the SFN questionnaire
In the first phase, three existing PSQs (the COPS, CQI-QOHC and the Patient Assessment of Chronic Illness Care (PACIC) questionnaires) 21,22,24 and a number of relevant items of a locally developed PSQ of the endoscopy department at the Maastricht UMC+ were selected for creating the conceptual content of the SFN-PSQ.
The COPS is a reliable questionnaire (α = .70-0.91) and consists of six relevant domains and 54 questions for disease-and treatment-related elements of care 24 : (1) Hospital admittance, (2) Nursing care, (3) Medical care, (4) Information, (5) Autonomy and (6) Discharge, with 5-point Likerttype answering satisfaction categories (1: unsatisfied to 5: very satisfied), and a patient characteristic on the general health status (with a 5-point Likert-type answering category (1: bad to 5: excellent). 24e CQI-QOHC questionnaire is based on eight important domains of the OC and/or function departments: (1) Hospital accessibility, (2) Welcome, (3) Facilities, (4) Welcome and stay, (5)   Respect and treatment by the doctor, (6) Communication, (7) Going home/discharge and (8) Overall quality assessment, consisting of 55 questions with different answer options (nine closed questions, nine multiple-choice questions, one open-ended question and 36 questions with a 4-5-point Likert scale). 25The CQI-QOHC has good reliability (α > .70),except for hospital accessibility (α = .64). 25 The PACIC questionnaire, based on the Chronic Care Model, has five predefined domains: (1) Patient activation, (2) Delivery system/ practice design, (3) Goal-setting/tailoring, (4) Problem-solving/ contextual and (5) Follow-up/coordination. 28 The PACIC is a 20item questionnaire, which used a 5-point response scale (from 1: almost never to 5: almost always). 22The domains of the PACIC have reliable Cronbach's α (from .71 to .83). 28e PSQ from the endoscopy department of the Maastricht The first step was to select the domains that reflected the multidimensional concept of patient satisfaction with the logistic process of the SFN service (Table 1).

| Phase 1b: Content validity and feasibility of the SFN questionnaire by the SFN expert team
Subsequently, the content validity and feasibility of the SFN-PSQ were further developed using a Delphi method with an expert SFN team. 29,30This team consisted of three neurologists, two medical doctors, two NPs and three administrative and/or logistic assistants.
A Delphi method is a method for obtaining and clarifying group judgements and provides a validated scientific method for resolving complex problems through expert consensus. 31,32The SFN-PSQ was sent by email to all team members for correction and emendation regarding the patient satisfaction content and logistic process of the diagnostic SFN service.Assessment of the SFN-PSQ was conducted by scoring (yes/no) on understandable language, relevance and redundancy for each question.The SFN expert team could also add additional relevant questions and could give qualitative comments on each domain.The content validity of the second concept of the SFN-PSQ was tested using the three-step test-interview (TSTI) method 33 with individual cognitive interviews.A basic definition of cognitive interviewing is administering draft survey questions while collecting additional verbal information about the survey responses, which is used to evaluate the quality of the response or to help determine whether the question is generating the information that its author intends. 34,357][38] The sampling method of key informants (based on sex, age, education level and hospital psychiatry visit) was used to select a variety of patients. 39,40

| Study procedure of the individual cognitive interviews of phase 2
With the first patient, a pilot interview was performed by two researchers (M.G. and M. E. J. B. G.) to test the interview procedure, interview guide and interview probes.Data from this pilot interview were also used in the analysis.All other consecutive interviews were performed by one researcher (M.G.), trained in interview skills.All patients were interviewed and audio-recorded once for single use for the study, and the interviewer's impressions of the interview process were documented by observation notes during the interviews.The three steps of the TSTI method 41 include an observation of response behaviour 42 and simultaneous verbalization (e.g., think-aloud), then exploring the formulations about the items and domains of the questionnaire that are difficult to understand (e.g., probing) and finally obtaining the thoughts and opinions of the patients. 41ink-aloud when completing a questionnaire is unusual and requires a well-founded explanation and some practice to elicit sufficient think-aloud behaviour. 33Therefore, as a warmup exercise, patients were asked to read aloud the first item (e.g., the date of the neurological analysis) or element (e.g., questionnaire instructions) in the SFN-PSQ and express any thoughts that came to their mind.To obtain more in-depth information about the thought process, probing questions were asked while the patient was thinking aloud and completing the SFN-PSQ.A probing question example for identification of response problems in concordance with the TSTI method is: 'Can you suggest any changes that would improve the interpretation of the questions in the SFN-PSQ?' Within each step of the TSTI, the SFN-PSQ was adjusted according to the patient's comments, which resulted in a third and final version of the SFN-PSQ.

| Data analysis
The COREQ (Consolidated Criteria for Reporting Qualitative Research) checklist was used to report the study. 43The results of the Delphi round were analysed, compared and discussed until consensus was reached by two researchers (M.G. and M.G.O.).Data collection and analysis of the individual cognitive interviews were continued until analytical data saturation was reached. 40,44A code system according to a classification by Willis and Lessler, 45 based on the Question Appraisal System, was used to classify the problems in answering the SFN-PSQ.In this method, the identified problems of the SFN-PSQ are classified by the interviewer/ researcher into one or more of the following seven categories: Clarity, Knowledge, Assumptions, Response Categories, Sensitively, Instructions and Formatting. 45Items and improvements that emerged during the sequential analysis were added to the SFN-PSQ and verified in the subsequent interviews.It was specifically examined whether the improvements solved the previously found problems when completing the SFN-PSQ.
Descriptive statistics were used to describe the sociodemographic and logistic features as well as the responses on question level the SFN-PSQ.

| Content validity by individual cognitive interviews with patients (phase 2)
From July 2019 till March 2020, nine consecutive patients participated for testing the content validity and feasibility of the second concept of the SFN-PSQ.Neurological analysis at the NDCU occurred in eight patients; in one patient, the neurological analysis was carried out at the OC.
General health status was rated as good in five patients.The median scores of anxiety and depression symptoms were 6.6 (confidence interval [CI]: 4.9-8.3)for anxiety and 6.4 (CI: 4.3-8.5)for depression.
The average duration for completing the SFN-PSQ was 14 min (IQR: 11-20).Furthermore, a differentiation was made for the logistic feature of the diagnostic SFN service (i.e., the NDCU or the OC), which was integrated into the final SFN-PSQ.

| Changes in the final questionnaire
Table 3 presents  During the Delphi round, the expert SFN caregivers achieved uniformity in the items about personal attention, expertise and privacy in one-on-one patient contacts.
After the input of the experts, the content validity of the SFN-PSQ was tested in a sample of patients with SFN-related complaints in daily practice.The most significant changes were adding domains and items concerning the waiting list period, diagnostic services and a consultation by the Hospital Psychiatry Care Provider.Also, a differentiation was made for both inpatient and outpatient neurologic analysis, ensuring that SFN-PSQ is applicable in all Furthermore, the individual cognitive interviews improved the clarity and intelligibility of the items, which has increased the comprehension of the SFN-PSQ from the patients' perspective.
Ultimately, this resulted in a feasible and reliable SFN-PSQ consisting of ten domains with 51 items, with a shorter average completion time than previously expected (approximately 20 min).This has shown that the final SFN-PSQ is suitable for measuring patient satisfaction in all patients who are visiting hospitals, where a similar diagnostic SFN service has been set up.
The strength of the study is that the content development of the SFN-PSQ was carried out in close collaboration with expert SFN caregivers with different backgrounds, but above all, the considerable influence of the patients on the content validity and user-friendliness of the SFN-PSQ.
Another strength is that the content validity of the questionnaire with experts and individual cognitive patient interviews confirmed that the SFN-PSQ is relevant and comprehensive.
A potential limitation is that the development and content validity of the SFN-PSQ were performed in a single-centre study and were not tested in centres outside the Netherlands.As there is only one diagnostic SFN service in the Netherlands, some domains and/or items of the SFN-PSQ may not be applicable as they are not consistent with the diagnostic SFN service in other countries.In addition, to ensure that translation is appropriate for the target group, 46,47 a linguistic validation procedure for the three SFN-PSQs has been performed.
Another limitation is that perhaps a small sample of mainly Caucasian expert SFN caregivers (85.7%) and patients (100%) participated in this study.However, the study was conducted according to the qualitative research guidelines and the interviews continued until data saturation was achieved.In addition, the SFN experts were from different professional backgrounds, ensuring that the content of the questionnaire was evaluated from different perspectives.
Second, although the number of individual cognitive patient interviews was small, the heterogeneity (sex, age, education, HADS scores) of the patients included is consistent with the general SFN population, 8,48 and the results may therefore be generalized to the general SFN population.Nevertheless, measurement of the crosscultural validity of the SFN-PSQ is recommended in a future study with a mixed-ethnicity study population. 49After the cross-cultural validation process of the SFN-PSQ, the questionnaire could be used as a model for other chronic conditions, such as Parkinson's disease or epilepsy, but also in developing care pathways for other medical specialties.

| CONCLUSION
The content validity and feasibility of the SFN-PSQ were developed and validated through item generation, expert opinions and interviews by patients.For future use of the SFN-PSQ, its construction based on the logistic and diagnostic SFN pathway could be a blueprint for other hospitals to measure patient satisfaction to optimize the quality of the logistic and diagnostic SFN pathway, other chronic diseases and care pathways from patients' perspective.
In 2009, an evidence-based standardized diagnostic SFN service was implemented at the SFN Center of the Maastricht University Medical Center+ in Maastricht (Maastricht UMC+), the Netherlands.The Department of Neurology is an independent specialty of the Brain and Nerve Center, which includes five other medical specialties, namely, Neurosurgery, Psychiatry, Psychology, Clinical Neurophysiology and Otorhinolaryngology.The diagnostic SFN service includes a neurological analysis with reserved slots for interview, examination and diagnostic tests at the Neurology Day Care Unit (NDCU) or the outpatient clinic (OC).
This study consisted of two phases.The first phase included a descriptive qualitative Delphi study with item generation for the content development and expert opinion of patient satisfaction related to the diagnostic SFN service.The second phase consisted of individual cognitive interviews with patients to test the content validity of the questionnaire.The study was performed at the Maastricht UMC+ in Maastricht, the Netherlands, and covered a period from February 2019 to July 2020.

( 5 )
UMC+ based on the CQI-QOHC questionnaire consists of seven domains: (1) Prior to Your Visit, (2) Welcome, (3) Respect and Treatment by the Doctor, (4) Respect and Treatment by the Nurse, Course of the examination, (6) Follow-up Services with 23 questions with a 4-point Likert scale (from 1: no, not at all to 4: yes, completely, and an extra answer option: not applicable) and (7) Overall Quality Assessment with a 1-to-10 rating scale.

T A B L E 1
Content selection of the four Patient Satisfaction Questionnaires for the first concept of the SFN-PSQ.
Patients 18 years of age and older, who completed the neurological analysis of the SFN diagnostic service in the last 6 months (on behalf of the recall period) and were living nearby the region of the Maastricht UMC+, were asked by phone to participate in an individual cognitive interview at the hospital for validating the second concept of the SFN-PSQ.The duration of the individual cognitive interview was approximately 60-75 min.The audio recordings were typed out by a research assistant and the content was analysed and categorized by one researcher (M.G.).

(
A) Prior to Your Stay at the NDCU/Waiting List Period and (B) Reception and Stay at the NDCU.The third question about the patient's ability to participate in treatment was replaced by another question from the PACIC (Patient activation: Asked about the patient's ideas when making a treatment plan).

3. 2 |From 5
Questionnaire adjustments with a Delphi method April to 29 April 2019, seven out of 10 expert team members (i.e., three neurologists, two medical doctors and two NPs) who were substantively involved in the SFN diagnostic service responded and discussed the content and feasibility of the SFN-PSQ.Three administrative and/or logistic assistants did not respond without specific reasons.The SFN expert team made a number of adjustments to the first concept SFN-PSQ, shown in Figure1.In summary, the corrections included more specific domains to measure patient satisfaction, grammatical changes to provide more clarity and deletion of items that showed overlap with other items or were not applicable.The results of this Delphi round were integrated into a second concept of the SFN-PSQ.In the final phase, the content validity and feasibility of the second concept SFN-PSQ were tested by patients.

Figure 2
Figure 2 shows the iterative process for adjusting the SFN-PSQ by the individual cognitive interviews.After the first interview, the SFN-PSQ was modified and presented in the interview with patient B. The results of the following two interviews in the third analysis round were combined and implemented, and the adjusted SFN-PSQ was presented to patients E and F. The final and third draft of the SFN-PSQ was created after the fourth analysis round of the last three patients (G, H and I).The most important changes to the SFN-PSQ questionnaire were addition of domains and items concerning the Prior to your Stay at the NDCU/Waiting List Period, the diagnostic services and consultation with the Doctor/NP of the Psychiatry Department.

Table 1
Core Questionnaire for the assessment of Patient Satisfaction Questionnaire; CQI, Consumer Quality Index; ECG, electrocardiogram; MD, medical doctor; NCS, nerve conduction studies; NDCU, neurology day care unit; NP, nurse practitioner; PACIC, Patient Assessment of Chronic Illness Care questionnaire; PSQ, Patient Satisfaction Questionnaire; Q, question; SFN, small fibre neuropathy.
presents the selection of the domains and items of the four PSQs.Autonomy as a separate domain (as in the COPS) was not considered suitable for a 1-day stay setting.However, two questions of the Autonomy domain were integrated into the SFN-PSQ domains:

Table 2
Categorization of problems of the cognitive individual interviews.