Abstract
Introduction
Despite the large and still growing range of treatment options available for gastroesophageal reflux disease (GERD), a significant proportion of patients receiving treatment for this chronic condition remain troubled by their symptoms. This has been shown to cause disruption to their daily lives and has an adverse impact on their quality of life [Gisbert et al. 2009a, 2009b; Jones et al. 2007]. The socioeconomic cost of GERD is considerable, as demonstrated by studies of GERD-related lost productivity, use of healthcare resources and the costs of both prescribed and over-the-counter (OTC) medications [Jones et al. 2007; Wiklund, 2004]. Moreover, a high proportion (39%) of patients have the more severe form of disrupting GERD, which is associated with significantly poorer quality of life, reduced productivity and increased healthcare resource utilization [Toghanian et al. 2010].
Pharmacological treatment is an essential factor in the management of GERD, but it is not the only one that relates to patient outcomes. The other factors that should be taken into account in clinical practice include the accuracy of diagnosis in primary care, measurement of patient needs as related to the intensity of symptoms and their impact on daily life, and the follow-up plan selected to assess response to the chosen therapeutic strategy. In addition, GERD treatment guidelines and restrictions on reimbursement of drug costs exist in many countries and there are strong national traditions within the management of GERD.
Therefore, we designed the GERD Management Project (GMP) to evaluate the effectiveness of a structured management approach to GERD compared with usual care using pooled individual patient data from five related cluster-randomized trials conducted in Europe. The trials were conducted in Austria, Spain, Italy, Norway and Sweden. In the structured treatment approach, patients were stratified according to symptom severity and treated accordingly, with potent acid suppression reserved for the group with high symptom load. The hypothesis is that such an approach would achieve a better clinical outcome than that achieved through usual clinical practice, but there was sufficient flexibility within the protocol to allow for country-specific adaptations, such that the results will also have meaning within an individual country. By pooling the data from five related studies, we increase the sample size and power to determine the efficacy of the structured approach to GERD management and we will also be able to evaluate the generalizability of study findings across countries.
We describe here the common rationale and design of the studies being conducted in each participating country. Country-specific features of the study designs are presented in Appendix 1.
GMP rationale
The five studies being integrated into GMP (ClinicalTrials.gov identifier: NCT00842387) compare a new management strategy for primary care patients with GERD with existing treatment practice to determine whether clinical outcomes are improved. The new strategy is based upon the use of the self-administered GERD Questionnaire (GerdQ) [Jones et al. 2009], a symptom-based diagnosis and management tool for patients with GERD. This validated tool was constructed following development of ‘the Montreal definition’, a patient-centred consensus definition of GERD based on symptoms [Vakil et al. 2006]. It has been shown to provide a diagnosis of GERD with an accuracy equal to that of a gastroenterologist and thus enables primary care physicians (PCPs) to diagnose and manage GERD with a lower need for referral or endoscopy [Jones et al. 2009]. A cut-off score of eight separates patients with a high likelihood of having GERD from those with low, or no likelihood of GERD.
Moreover, use of GerdQ enables patients to be stratified according to the frequency and impact of their symptoms and thus to be treated optimally [Jones et al. 2009]. This should both improve patients’ health-related quality of life and avoid unnecessary investigation and under- or over-treatment. GerdQ can also be used to follow and measure patients’ response to treatment over time.
Because the healthcare systems of the participating countries have different national guidelines for the management of GERD, each study included in GMP was given the flexibility to modify the standard study protocol as needed, in collaboration with local regulatory and healthcare authorities and participating physicians. For example, in Spain and Sweden, a local preference for use of the Reflux Disease Questionnaire (RDQ) tool rather than GerdQ for evaluation of the respective primary endpoints was accommodated. The local modifications incorporated ‘real-world’ clinical practices in that locality and should facilitate the implementation of the strategy to future clinical practice with a minimum of further refinement.
Features of GMP design by country
Key variations from standard protocol.
EQ-5D, EuroQol-5D Questionnaire; GERD, gastroesophageal reflux disease; HRQL, health-related quality of life; RDQ, Reflux Disease Questionnaire; WPAI-GERD, Work Productivity and Activity Impairment Questionnaire for patients with GERD.
In Austria and Spain, the studies were considered noninterventional by regulatory authorities, and were cluster-randomized for the purpose of assessing the effect of the structured clinical pathway on GERD patients. The implementation of the recommended structured pathway consisted of a specific training session, given by specialists, at a pool of primary care centres, although the decision on whether or not to follow the recommended pathway remained with the individual PCP. This allowed for an estimation of uptake of the recommended pathway, and an assessment of the factors influencing uptake.
In Italy, the study was considered interventional by regulatory authorities, and was run as a PCP level, cluster-randomized, controlled trial comparing a structured clinical pathway versus usual care in patients with GERD. The implementation consisted of training sessions on the new clinical pathway.
In Norway, the study was conducted as an evaluation of a symptom-based (GerdQ) versus endoscopic approach for the diagnosis, choice of treatment and evaluation of GERD, in which the new structured pathway in the diagnosis and treatment of GERD was compared with the standard clinical pathway.
In Sweden, the study was conducted as an evaluation of the new management strategy for GERD, in which the participating primary care centres were randomized (one to one) to implementation of the structured clinical pathway or to management of patients according to local clinical routines.
Patients
The patient population in all five studies was representative of primary care patients with symptoms suggestive of GERD, regardless of severity. Men and women aged at least 18 years and capable of understanding and completing the questionnaires were recruited, and informed consent obtained. Patients with alarm symptoms such as dysphagia/odynophagia, anorexia, anaemia, unintentional weight loss, abdominal mass or upper gastrointestinal bleeding were referred for specialist treatment and excluded from the studies. Patients were free to withdraw from the studies at any time, without this affecting their medical care or changing the therapeutic pathway through which they were managed.
Standard study protocol for local adaptation
A schematic representation of the standard study design is presented as Figure 1. Modifications to the standard study protocol were allowed to take account of national guidelines, and key modifications by country are described in Table 1. Key aspects from individual country-specific protocols are listed in Appendix 1.
Standard study flow chart.
The standard protocol for use in local adaptation, as required, was as follows: demographic and clinical information was collected for both patient groups (new strategy versus standard care) at the start of the study and with a follow-up visit after 4 weeks to collect efficacy data. Patients who had not improved sufficiently at 4 weeks were reassessed at 8 weeks.
Among primary care centres randomized to the new management strategy, implementation consisted of detailed explanation of the structured approach to physicians, who may use the approach to treat patients at their discretion. The physicians’ adherence to the structured clinical pathway was monitored. The physicians in the control groups were informed that the aim of the study was to determine the effect of treatment prescribed to typical GERD patients in usual clinical practice, and that the symptom profile of these patients were to be assessed through questionnaires. To maintain the integrity of randomization, the implementation and control groups did not include centres that were geographically close. Differences in the use of resources between centres that implemented the pathway and those that did not were monitored.
Patient assessments
Patient gender, age, weight, smoking status and alcohol intake were recorded at the study start. Any previous gastrointestinal diagnoses (dyspepsia, hiatus hernia, abdominal pain or peptic ulcer) were also documented.
Patients were classified into different groups according to their GerdQ score. A score of 7 or below indicates that the patient has a low probability of GERD, whereas a score of 8 or higher positions the patient as more likely to be a GERD patient [Jones et al. 2009]. Within these two disease levels, patients who had an impact score of 2 or less were classified in the low/moderate impact GERD group, whilst those with a score of 3 or more were classified in the high impact GERD group. Patients with a GerdQ impact score ≤2 were treated with generic proton-pump inhibitors according to local guidance, and patients with an impact score ≥3 were treated with esomeprazole 40 mg once daily.
Treatment response was determined by the GerdQ score as follows: patients should have a score of less than 2 for each of the items heartburn, regurgitation, need for OTC treatment and sleep disturbance, on at most 1 day during the previous 7 days.
Other outcomes measures in the original studies varied by country, but generally included: treatment failure, treatment satisfaction, measures of health status and work productivity, safety, and both direct and indirect costs associated with management processes such as physician, specialist and emergency visits, diagnostic tests, surgery and hospitalizations.
Statistical analysis
A pooled analysis based on individual patient data is planned. Mixed regression models will be used to calculate an overall effect estimate (implementation of the structured approach versus usual clinical care), taking account of clustering by centre, and adjusted according to baseline differences between intervention and control centres. An analysis stratified by country is also planned. If significant interactions by country are identified, the results for each country will be presented separately. The statistical analysis will be performed using Stata 11 software (StataCorp LP, College Station, TX, USA).
Discussion
We have described here the rationale and design of a project to assess the utility of a new management approach for GERD in primary care, which encompasses major between-country differences in healthcare systems, management approaches and physician preferences. The project was undertaken as an acknowledgement of the importance of evaluating management protocols of patients/diseases in primary care. This is an area that is perhaps under-researched, not only in relation to the evaluation process itself but also in terms of the resultant clinical outcomes.
This project evolved through a unique consultation approach and, in each country, utilized the protocols that were developed and agreed upon following liaison with local healthcare authorities and physicians. This consultative and modifiable approach is a major strength of the project, as it will allow results that are specific to each participating country to be obtained. This aspect will determine whether the GerdQ management tool has real clinical applicability within each country, despite differences in local practice. Evaluation of a new management tool is a complex task and achieving this across a range of regulatory backgrounds and differing national guidelines is a significant achievement. The studies are currently nearing completion or have been completed in each country.
In addition to evaluation of the GerdQ management tool, key data related to disease management costs (new management pathway versus usual care) will be generated. A critical factor in assessing the viability of the introduction of the process/protocol is the impact on disease management-related costs. Provision of healthcare costs associated with potential implementation of initiatives intended to improve patient care is a critical requirement for informed decision-making in healthcare administration.
The contribution of a symptom-driven approach, such as the GerdQ management tool, to improved patient outcomes will be evaluated in the context of other management strategies for GERD, such as treatment decisions based largely on endoscopic findings (as currently employed in Norway) and step-up/step-down approaches [Howden et al. 2001; Inadomi et al. 2001]. It is important to note in this context that the majority of GERD management guidelines support symptom-driven care without recourse to endoscopy for patients with uncomplicated disease (i.e. in the absence of alarm symptoms) [Kahrilas et al. 2008; Armstrong et al. 2005; DeVault and Castell, 2005].
The results of this project should demonstrate whether or not a locally adapted management strategy for GERD in primary care, based upon GerdQ as a tool in the diagnosis and evaluation of therapy tailored to the needs of the individual patient, is beneficial compared with usual care. The project will therefore provide valuable information on the suitability of GerdQ as a gold standard in the objective diagnosis and monitoring of patients with disrupting GERD, along with associated disease management costs in relation to usual care.
In summary, if the results show that this management strategy can be successfully adapted to local regulatory procedures and both national and local GERD management guidelines, then the new strategy could be implemented into local clinical practice with minimal further adjustment if it proves to be superior to standard practice.
Footnotes
Funding
This study was supported by AstraZeneca, the manufacturer of esomeprazole. We thank Jackie Campbell, from inScience Communications (a Wolters Kluwer business), who provided medical writing support funded by AstraZeneca.
Conflict of interest statement
Authors employed by AstraZeneca (MT and JN) participated in study design and management. JP, VG, LA, ET, MG, EG and JGH had full access to all data and had final responsibility for the decision to submit for publication.