Health-Related Quality of Life and Other Patient-Reported Outcomes in the European Centralized Drug Regulatory Process: A Review of Guidance Documents and Performed Authorizations of Medicinal Products 1995 to 2003

Authors


Ágota Szende, Covance Health Economics and Outcomes Services, Springfield House, Hyde Street, Leeds LS2 9LH, UK. E-mail: szende@covance.com

ABSTRACT

Objectives:  The objective of this study was to review and analyze the use of health-related quality of life (HRQL) and other patient-reported outcome (PRO) evaluations for the approval of new pharmaceutical products by the European Medicines Agency (EMEA).

Methods:  All published EMEA guidance documents and regulatory information for products authorized at the EMEA and appearing in the European Public Assessment Report (EPAR) database between 1995 and 2003 were examined for reference to HRQL and other PROs.

Results:  More than half of the guidance documents for clinical investigation of pharmaceutical products in specific disease areas included reference to HRQL or other PROs. Guidance notes for 10 conditions indicated PROs can serve as primary endpoints in clinical trials, among which three included HRQL outcomes. The review of EPAR documentation uncovered HRQL and other PRO data for 34% of the drugs registered during the period of the review, with cancer-related treatments most frequently including PRO data. There was a trend toward increasing HRQL and other PRO claims in regulatory documents of pharmaceutical products in recent years, with the proportion exceeding 30% from 1999 to 2003.

Conclusions:  There is further scope for health outcomes researchers and regulatory decision-makers to contribute to the more efficient utilization of PROs and HRQL outcomes. Health researchers need to better justify the inclusion of these outcomes in clinical trials and highlight the added value of PRO data; while the regulators should develop harmonized procedures and capacities to adequately appraise the submitted information.

Introduction

In the European Union (EU), both the European Commission (EC) and the national health authorities of the EU member states have the authority to register medicinal products. In an effort to harmonize the review process across member states, the EC established a formal, centralized authorization procedure in 1995, delegating its certifying power to the European Medicines Agency (EMEA). The EMEA consists of members or delegates from the participating national health authorities and serves as a coordinating body, bringing national experts together for the centralized evaluation process. Within the EMEA, the pan-European Committee for Medicinal Products for Human Use (CHMP) is responsible for producing guidelines and making decisions about the authorization of most medicinal products for human use (Before 2004, it was called as the Committee for Proprietary Medicinal Products—CPMP). The CHMP's work is, in turn, supported by several working parties, such as the Efficacy Working Party (EWP). With the increasing emphasis on harmonization across the EU and ongoing interest in placing all drug approvals through a centralized review process [1], the importance of the EMEA continues to increase.

The EMEA was established at a time when health-related quality of life (HRQL) and other types of patient-reported outcomes (PRO) were emerging as important components of the evaluation of new pharmaceutical products worldwide. Although scientific evidence had been accumulating on the usefulness of validated measures of self-reported health outcomes in evaluating a broad range of health care interventions, relatively little was known about how this type of information was incorporated into the drug regulatory decision-making process [2,3]. For this reason, the EMEA began the authorization process without a defined approach for evaluating HRQL and other PRO data.

Despite the absence of specific guidance documents on HRQL and other PROs, and of any demand from the EMEA, pharmaceutical companies have elected to include these outcomes in clinical trials and submit these data to the EMEA as part of the review process. The first case where HRQL data appeared in product regulatory documents was docetaxel, one of three products registered at the EMEA during its first year of operation. Docetaxel is indicated for the treatment of locally advanced or metastatic breast cancer. Evidence for impact of the drug on HRQL, measured by the European Organization for Research on Treatment of Cancer (EORTC) quality of life questionnaire, was provided from a large phase III study in which docetaxel was used in combination with doxorubicin. HRQL results were discussed both under efficacy and safety sections of the Scientific Discussion document. In terms of efficacy, the document stated that no difference could be observed in HRQL between the two patient groups, indicating no adverse effects of treatment. With respect to safety considerations, the document indicated that the stability of HRQL during treatment reflected that drug toxicity was manageable in both treatment arms. These findings also appeared in the summary of product characteristics (SPC) document, stating, “In both arms, quality of life measured by the EORTC questionnaire was comparable and stable during treatment and follow-up.” This early experience with HRQL in the authorization process serves as a baseline for a review of the use of HRQL in regulatory submissions in Europe between 1995 and 2003.

A previous review by Apolone et al. [4], examined the EMEA's recommendations on the use of HRQL data in clinical trials, and found that HRQL was discussed as a potential efficacy endpoint in several of the guidelines. The authors concluded that the recommendations were vague and inconsistent across disease areas. The objective of this study was to update the information on HRQL and expand the work by providing an in-depth review and analysis of HRQL and other PRO information in EMEA guidance documents and authorizations of medicinal products performed between 1995 and 2003. This review forms the basis for recommendations on future efforts to bridge health outcomes research and European drug regulatory decision-making.

Methods

This study reviewed the role of HRQL outcomes and other PROs in the European centralized regulatory process. Although there is no one single definition of PROs accepted by all researchers, broad agreement exists on a number of its key characteristics. We adopted a recently published definition of PROs as recommended by an international harmonization task force [5]. According to this definition PROs represent the patient's report of a health condition and its treatment. PROs are inherently subjective and they provide a unique contribution to the drug development process. HRQL is one of the several types of PROs and represents the patient's evaluation of the impact of a health condition and its treatment on daily life, well-being and functioning. HRQL is of multidimensional nature, with type and number of dimensions varying across diseases. Other PROs include patient-reported symptoms, preference, satisfaction with treatment, functional status, disability, and other outcomes.

During the review, the above defined broad definition of PROs was applied to identify products for detailed review. Patient-reported symptoms were included only if they were reported to be collected in a systematic way, that is, from each subject using standardized procedures during which the patient rated his or her health status. When relevant, we also included outcomes reported by proxy respondents. The type of PRO response scale or the number of items used in the PRO evaluation was not an inclusion or exclusion criteria. Those products for which the information in the regulatory document was insufficient to decide whether a PRO was actually measured were not counted among those with PRO data.

Search Strategy

A systematic manual search was performed on guidance documents and product level regulatory documents published on the EMEA's Web site. The time frame of the review was the period between 1995 and the end of 2003. The documents were read individually to identify all PROs. Electronic searches on selected PROs (i.e., quality of life, health status, disability, well-being, patient satisfaction) served to check the manual review. Given the varied use of different terms for PROs in EMEA guidance documents, we thought that the manual review would provide the most comprehensive information.

Systematic searches were performed on publicly available official regulatory documents at two main levels: (1) general and disease-specific guidance documents; and (2) product-level regulatory documents.

Guidance documents.  EMEA/CHMP disease/drug-related guidance documents provide recommendations for submissions related to a particular disease (e.g., acute stroke) or medicinal group (e.g., antiarrhythmics). These documents are normally prepared by the Efficacy Working Party and are then reviewed and endorsed by the CHMP. There are three main types of guidance documents at this level: Concept papers, Points to consider, and Guidelines. These documents are hierarchical in terms of their developmental status, the scientific evidence on which they are based, and the expectations for compliance. Thus, recommendations included in the guidelines have the strongest impact on clinical trials while those included in the concept papers have the lowest impact. The status and the content of guidance documents change over time through development and updating.

For this study, all general and disease/drug-related guidance documents on regulatory clinical trials developed through the EMEA's Committee for Proprietary Medicinal Products with mention of HRQL and PRO data were identified. General guidance documents are relevant to drugs across all diseases while disease/drug-related guidance documents on clinical trials provide recommendations specific or unique to the stated disease or therapeutic indication or groups of medicinal products.

Health-related quality of life and other PRO data were summarized along several dimensions: the type of HRQL or other PRO data mentioned in the recommendations, whether these data were recommended as primary or secondary endpoints, and whether there were any specific instruments mentioned or recommended for use in clinical studies.

Product-level documents.  The review also included a search for HRQL and other PRO evaluation data in product-level regulatory documents. The European Public Assessment Report (EPAR), the published documentation of product-level regulatory information published by the EMEA, was reviewed for all product-level regulatory information for pharmaceutical products registered with the EMEA between 1995 and 2003. Four types of product-related EPAR documents were examined:

  • • Scientific Discussions: a detailed summary of the reviewed material submitted by the manufacturers to the EMEA to obtain marketing authorization for that product.
  • • Summary of Product Characteristics: a description of the product for the information of physicians as published in national drug compendia.
  • • Package Leaflets: information intended for the patient.
  • • Abstracts: a one-page document that summarizes the key findings of the evaluation process and the decision on granting authorization for the general public.

To interpret the results of this component of the review, it is important to understand the legal and regulatory background on the intended content and use of these documents.

The Scientific Discussion is the most detailed description of the scientific evidence from the regulatory evaluation of each new product. The Scientific Discussion contains commentary made by the CHMP. The purpose of the Scientific Discussion is to achieve transparency of the decision-making process and to describe to the public how decisions were made about granting a marketing authorization for a new pharmaceutical product.

As outlined in the European Commission's Guideline on Summary of Product Characteristics [4], the objective of the SPC is to provide “information for health professionals on how to use the medicinal product safely and effectively.” After approval is granted by the EMEA, the SPC serves as the description of the product in national drug compendia. According to the guidance [6] the SPC is to include specific pharmacologic and manufacturer-related information. Clinical efficacy can only be included in the “pharmacodynamic properties” section of the SPC but generally “no information is expected.” The guideline, however, states that in some cases, such as in new therapeutic areas, “main results (statistically compelling) . . . could be mentioned here in condensed form.” This is compulsory for medicinal products approved under exceptional circumstances.

The Package Leaflet contains information intended for the patient. This summary information is provided to the patient on obtaining the medication. According to the official requirements [7,8], the package leaflet shall be based on information provided by the SPC. As a general approach, no efficacy data are to be included in the package leaflet. Nevertheless, in accordance with Directive 92/27/EEC [8], the package leaflet may include “. . . other information compatible with the summary of product characteristics, which is useful for health education, to the exclusion of any element of a promotional nature.”

A systematic search was performed on the Scientific Discussion of each product registered at the EMEA between 1995 and 2003, using the EPAR documents and the search terms listed above. The purpose of this phase of the review was to determine the extent to which HRQL and other PRO data were considered during the decision-making process, as reflected in the Scientific Discussion document. It is important to note that information presented in the Scientific Discussion reflects data that were considered or accepted for consideration during the decision-making process, rather than all of the information submitted for review. It is possible therefore that HRQL or other PRO data were submitted, but do not appear in the Scientific Discussion of the product.

If HRQL and other PRO data were reported in the Scientific Discussion, a further review was performed on the SPC, the package leaflet, and the abstract to determine how the HRQL or PROs were measured, presented, interpreted, and commented. Although, based on the above described regulations the content of the SPC and the package leaflet should not normally include any efficacy data, it was still interesting to study how frequently and in what ways HRQL is and other PROs are presented in these documents.

Validation

Results of the systematic review of published regulatory documents were validated through semistructured interview with individuals employed by the EMEA. The interview centered around five main topics: conceptual definition; decision-making process, organization, and the role and place of HRQL data; currently published official EMEA/CHMP recommendations on HRQL data and their impact on submissions; comparisons with the Food and Drug Administration (FDA); and opinion, future plans. The objective of the semistructured interview was to interpret and validate the findings of the systematic review of regulatory documents. All information gained through interview is presented as a summary and the names of the participants will remain confidential.

Numeric Summary

To assess trends over time, the number of registered products with HRQL and other PRO data were summarized by year.

Results

Guidance Documents

General.  During the time period of this review (1995–2003), no general guidance document was published by the EMEA on HRQL or other PRO evaluation in clinical trials performed for regulatory purposes.

One general guidance document was published that included recommendations on HRQL measurement: the International Conference on Harmonization (ICH) guidance document on statistical principles for clinical trials [9,10]. This document suggests that the primary clinical endpoint in any trial be the variable capable of providing the most clinically relevant and convincing evidence directly related to the primary objective of the study, which is generally to evaluate treatment efficacy. Safety/tolerability may serve as the primary variable, and will always be an important consideration. The guidance document also included the following comment: “measurement related to quality of life and health economics are further potential primary variables.” No recommendations were provided specific to the measurement of HRQL. In general, the guidance recommended that the selection of the primary variable reflect the accepted norms and standards in the relevant field of research. There should be sufficient evidence that the primary variable can provide a valid and reliable measure of some clinically relevant and important treatment benefit in the patient population described by the inclusion and exclusion criteria.

Semistructured interviews with representatives of the EMEA not only confirmed that there were no general guidance documents on HRQL or PROs, but indicated that the EMEA had not adopted any explicit definitions of HRQL for use during the review process, because of the perceived lack of a consensus definition for HRQL in the scientific community. Accordingly, the EMEA accepts definitions of HRQL on a submission-by-submission basis. The primary concern of the EMEA is the use of solid scientific methodology and robustness with each individual submission. Research should be designed to demonstrate that drugs are safe, efficacious, and of good quality. The selection of any outcomes, including HRQL and PROs, should be justified in these terms; the design and implementation of the study should be sound.

Disease-specific guidance documents. Table 1 summarizes the inclusion of HRQL and other PRO data in disease/drug-related guidance documents issued by the EMEA from 1995 to 2003. Of the 57 documents issued, more than half (53%) included recommendations or statements on the use of HRQL and other PROs in clinical investigations of medicinal products. Four of the identified guidance documents (i.e., sepsis, social anxiety disorder, antiarrhythmics, and acute cardiac failure) suggested that HRQL is an important aspect of the disease, treatment or side effects, but did not contain any explicit recommendations on the use of such measurement in clinical evaluations. An additional review of guidance documents on biotechnology and blood products revealed only one document referring to HRQL as relevant aspect of treatment [11], although had no specific recommendations on its evaluation.

Table 1.  HRQL and other PRO data in guidance documents published by the European Agency for the Evaluation of Medicinal Products (EMEA), 1995–2003
DocumentHRQL and other PROs includedYearHRQL and other PROs not includedYear
  • *

    HRQL cited as an important aspect of the disease or treatment, but no explicit recommendations are made on the inclusion of HRQL measurement in clinical trials.

  • Year of first version issued.

Concept papersSepsis*2003Ankylosing spondylitis2003
Social anxiety disorder*2003Chemotherapy-induced nausea and vomiting2003
  Chronic hepatitis B2003
  Dyslipoproteinemia2001
  Peptides and proteins2002
Points to considerAcute stroke1998Acute coronary syndrome1998
Amyotrophic lateral sclerosis1998Acute respiratory distress syndrome1997
COPD1998Antifungal agents2001
Crohn's disease1999Diagnostic agents1998
Irritable bowel syndrome1997Hemoatopoietic growth factors1999
Neuropathic pain2003Hormone replacement therapy1997
Osteoarthritis1997Prophylaxis of intra- and postoperative venous thromboembolic risk1998
Rheumatoid arthritis1995  
Guidelines— draftAnxiety disorder (draft)2002Impaired hepatic function (draft)2002
Allergic Rhinoconjunctivitis (draft)2002  
Obsessive compulsive disorder draft)2002  
Panic disorder (draft)2002  
Psoriasis (draft)2002  
Guidelines—adaptedAsthma2001Antibacterial medicinal products1995
Acute cardiac failure*2001Bipolar disorder1998
Alzheimer's disease1995Depression1997
Antiarrhythmics*1995Diabetes mellitus2000
Cancer1995Epileptic disorders1998
Cardiac failure2003Hypertension1995
HIV infection2002Impaired renal function2002
Migrain2001Lipid disorders2003
Multiple sclerosis1998New vaccines1997
Nociceptive pain2000Post menopausal osteoporosis in women1995
Parkinson's disease1995Schizophrenia2001
Peripheral arterial occlusive disease (PAOD)1998Steroid contraceptives in women1998
Urinary incontinence2001Thromboembolic disease1998
Stable angina pectoris1995Thrombosis in acute myocardial infarction2001
Weight control1996  
No. issued30 (53%)27 (47%)

Although slightly more than half of the guidance documents included statements or recommendations on HRQL or PRO data, almost half of the documents did not. There was no clear pattern in the difference between those diseases/drug groups for which recommendations included HRQL and other PROs data and those diseases which did not. In fact, many of the documents that did not include mention of HRQL data were in chronic diseases, such as diabetes, depression, and schizophrenia, where new treatments could have a substantial impact on patient HRQL. The impact of these chronic illnesses on HRQL has been demonstrated, and there are instruments available for assessing HRQL and other relevant outcomes in these diseases.

Interviews with the EMEA confirmed that, in the absence of a general guidance document on the use of HRQL and PROs or in the development of new guidelines, disease-group level review and recommendations are dependent on the experience of the EMEA and the views of the individual experts. In addition, HRQL or other PRO data may be submitted by the manufacturer and considered by the EMEA whether or not the relevant EMEA/CHMP recommendation explicitly mentions these endpoints. If HRQL or other PRO data are submitted for drugs where the relevant EMEA/CHMP recommendation does not explicitly mention these as possible endpoints, the EMEA still considers them. The importance of HRQL or other PRO data in the review and decision-making process is based on its relevance to a particular drug and study according to the judgment of the CHMP.

Discussion of HRQL and other PROs as efficacy endpoints.Table 2 presents a summary of key issues related to HRQL and other PRO data presented in the disease-group level guidance documents [11–38]. Because reference to sepsis, social anxiety disorder, antiarrhythmics, and acute cardiac failure did not include specific recommendations, these documents are excluded from the table. All but three of the guidelines addressed HRQL. In 15 cases, recommendations included PROs or specific HRQL domains, namely, patient-reported symptoms, discomfort, pain, disability, physical functioning, general well-being, and patient's global assessment of improvement. Most of the guidelines regarded HRQL or PROs data as potential efficacy endpoints but the guidance on HIV medication recommended that HRQL data can serve as an additional safety endpoint.

Table 2.  Health-related quality of life (HRQL) and other patient-reported outcomes in disease-group level guidance documents issued by the European Agency for the Evaluation of Medicinal Products (EMEA), 1995–2003
Disease groupRecommended primary
clinical endpoints
Recommended HRQL and
other PROs
Mention of example
instrument
Points to consider
Acute strokeSurvival; disability groupHRQLN/A
Amyotrophic lateral sclerosisSurvival; Muscle strength; Respiratory function; Functional test of disabilityHRQLN/A
COPDLung function; Symptomatic benefitHRQL; patient-reported symptomsSt. George's Respiratory Questionnaire, SGRQ
Crohn's diseaseProportion of patients achieving/maintaining remissionHRQLInflammatory Bowel Disease Questionnaire, IBDQ
Irritable bowel syndromePatient's global assessment of symptoms; Abdominal pain and discomfortHRQL; patient-reported symptoms, pain and discomfortN/A
Neuropathic painMultidimensional, patient-rated, pain assessmentHRQL; Patient-reported pain; Patients’ global assessment; Functional performance; Social performanceMcGill Pain Questionnaire (MPQ); MPQ Short Form (MPQ-SF); Neuropathic Pain Scale (NPS)
OsteoarthritisPain and functional disabilityHRQL; patient-reported pain and disabilityWestern Ontario McMaster University, WOMAC
Rheumatoid arthritisJoint count (swollen or painful); disease activity (assessed by the physician or patient); pain score (patient's assessment of ain, VAS); acute phase reactants, physical function and HRQL; radiographHRQL; patient-reported pain, physical functioningHealth Assessment Questionnaire (HAQ), Arthritis Impact Measurement Scale (AIMS), SF-36 Health Survey
Guidelines
Allergic Rhino-conjunctivitis (draft)Patient-rated symptomsPatient-rated symptomsN/A
Alzheimer diseaseCognition; ADL; Overall clinical responseHRQLN/A
AsthmaAirway obstruction; lung function; symptom-based clinical endpointsHRQLN/A
CancerTumor burden; disease stabilization/prolonged progression-free survival, and overall survival; Progression-free recurrence-free/relapse-free survival; Response rate; Symptom control/HRQLHRQLN/A
Cardiac failure (acute and chronic)Mortality; Clinical symptoms (chronic cardiac failure)HRQL; Self-assessed global clinical status; Patient-rated symptomatic improvements in dyspnea, fatigue, mental confusion, and general well-beingMinnesota Living with Heart Failure Questionnaire
General anxiety disorder (draft)Hamilton Anxiety Rating Scale (HAM-A)Self-assessed disability; HRQLSheenan Disability Scale
HIV infectionViral load and CD4+ T-cell countsHRQLN/A
MigrainePercent of patients pain-free at 2 h after administration of medicationHRQL; Global evaluation of medication by patientN/A
Multiple sclerosisDuration and severity of relapses (acute treatment); Disability progression (disease modifying treatment)HRQL; “Patient's global opinion”N/A
Obsessive compulsive disorderYale-Brown Obsessive Compulsive Scale (Y-BOCS)Global self-assessed scales to measure social and occupational functioningSheenan Disability Scale
PainSelf-assessed pain Multidimensional assessment toolsSelf-assessed pain Multidimensional assessment toolsVisual Analog Scale (VAS); Numerical Pain Scale (NPS); Pain Descriptor Scales (PDS); McGill Pain Questionnaire; Short Form McGill Pain Questionnaire; Minnesota Multiphasic Personality Inventory; Psychological Pain Inventory; McGill Comprehensive Pain Questionnaire; Pain Profile; Multidimensional Pain Inventory
Panic disorder (draft)Frequency and severity of attacks; Panic Disorder Severity Scale (PDSS); Panic and Agoraphobia Scale (PAS)Self-assessed disability; HRQLSheenan Disability Scale
Parkinson's diseaseSuccess and failure to treat symptoms measured by various rating scales depending on the treatment objectivesHRQLN/A
Peripheral arterial occlusive diseaseInitial and absolute claudication distance (stage I); pain at rest and response-based endpoints (stage II); ulcer healing and response-based endpoints (stage IV)HRQLN/A
Psoriasis (draft)Psoriasis area and severity index (PASI); visual assessment of index lesions, BSA measurement, total severity sign score (TSS); physicians’ global assessment (PGA) of all lesions on a 6- or 7-point scalePatient's assessment of global improvement (same as PGA but responded by patient); self- administered PASI (SAPASI); HRQL scales validated in dermatologyDermatology Life Quality Index (DLQI); Dermatology Quality of Life Scales (DQLS); Skindex; Psoriasis Disability Index (PDI), Psoriasis Life Stress Inventory (PLSI)
Stable angina pectorisExercise capacity; anginal pain; HRQL; morbidity and mortality; and ischemic pisodesHRQLN/A
Urinary incontinenceUrodynamic endpointsHRQL; Patient-rated symptomsN/A
Weight controlDemonstrated weight lossHRQLN/A

Ten (38%) of these guidance documents suggested that HRQL or other PROs can be considered either primary or secondary endpoints. These documents addressed submissions in the areas of allergic rhinoconjuctivitis, cancer, chronic obstructive pulmonary disease (COPD), irritable bowel syndrome (IBS), osteoarthritis, pain, Parkinson's disease, rheumatoid arthritis, stable angina pectoris, and neuropathic pain. Three of these guidance documents (i.e., rheumatoid arthritis, cancer, and stable angina pectoris) recommended HRQL among possible primary endpoints. The document on COPD advised that symptomatic benefit serve as a coprimary endpoint (with lung function) while HRQL serve as a secondary endpoint. The St George's Respiratory Questionnaire (SGRQ), which is commonly used as a condition-specific indicator of HRQL in this population, was mentioned as an example of a symptom outcome measure. Interviews with the EMEA pointed out that guidance documents offer recommendations, rather than mandates, leaving open the possibility that HRQL and other PRO outcomes can serve as a primary endpoint, provided that there is a strong rationale and that the instrument selection is justified.

Reference to specific HRQL and other PRO measures.  Most of the guidance documents did not specify whether generic or condition-specific instruments should be used to evaluate HRQL. The guidance document on peripheral arterial occlusive disease suggested that HRQL “be measured by properly validated generic and disease-specific questionnaires,” while the document on amyotrophic lateral sclerosis (ALS) indicated that if generic HRQL measures are used they should be validated in ALS patients. The need for valid disease-specific HRQL measures was also highlighted in the guidance documents related to Alzheimer's disease (AD). The guidelines for AD indicate that no specific tool can be recommended. The exact wording is as follows: “Although quality of life is an important dimension of the consequences of diseases, the lack of validation of its assessment in AD does not allow specific recommendations to be made as yet. When adequate instruments to assess this dimension in patients and their care givers become available, quality of life assessment may be justified in AD trials.”

Wherever an HRQL or another PRO measure was mentioned as an example in the guidance documents, it was always a condition-specific instrument. The only exception was the guidance on rheumatoid arthritis which mentioned the generic SF-36 Health Survey instrument as an example.

Only the guidelines on urinary incontinence addressed the issue of how HRQL information should appear in the SPC. This guidance document suggested that if clinically relevant changes were found in HRQL data, they should be included in the pharmacodynamics section of the SPC, the section generally containing efficacy data.

The interviews confirmed that instruments mentioned in disease-related EMEA/CHMP recommendations serve as examples only, should not be construed as an endorsement by the EMEA, and therefore do not preclude the use of other measures. Pharmaceutical industry investigators are encouraged to use well-developed and validated outcome measures. Any clinical trial with HRQL and PRO outcomes to be submitted for consideration by the EMEA should have prespecified hypotheses, well-designed study protocols, planned statistical analyses, and careful implementation.

Product-Level Information

Scientific discussion.  According to the EPAR documents, 237 products were registered by the EMEA from 1995 to 2003. The systematic search identified 81 (34% of those submitted) products that included HRQL or other PRO information (Table 3).

Table 3.  Pharmaceutical products registered with the European Agency for the Evaluation of Medicinal Products (EMEA) with health-related quality of life (HRQL) and other Patient-Reported Outcomes (r-PRO) in the scientific discussion, 1995–2003
ProductGeneric nameIndicationHRQL or other PRO instrument
  • *

    PRO data also is included in the SPC document.

Aerius*DesloratadineRhinitisSF-36 Health Survey, rhinoconjunctivitis QoL questionnaire
Aldurazyme*LaronidaseEnzyme replacement therapy in mucopolysaccharodosis IHealth Assessment Questionnaire (HAQ), Apnea Hypopnea Index (AHI)
Allex*DesloratadineRhinitisSF-36 Health Survey, rhinoconjunctivitis QoL questionnaire
AravaLeflunomideRheumatoid arthritisHealth Assessment Questionnaire
Azomyr*DesloratadineRhinitisSF-36 Health Survey, rhinoconjunctivitis QoL questionnaire
Bextra*ValdecoxibOstheoarthritis, rheumatoid arthritis, primary dysmenorrheaWestern Ontario and McMaster Universities (WOMAC) OA Pain Index and OA Physical Function Index, Patient's Global Assessment of Arthritis, Patient's Overall Assessment of Pain measured by Visual Analog Scale (VAS), Patient's Global Assessment of Disease Activity, Modified Health Assessment Questionnaire Disability Index (mHAQ)
CaelyxDoxorubicin HydrochlorideAIDS-related Kaposi's sarcomaMOS-HIV Health Survey
Cialis*TadalafilErectile dysfunctionSexual Encounter Profile, International Index of Erectile Dysfunction
ComtessEntacaponeParkinson's diseaseON/OFF period assessment based on patient symptom diary
ComtanEntacaponeParkinson's diseaseON/OFF period assessment based on patient symptom diary
CotronakRibavirinHepatitis CUnspecified HRQL measure
Depocyte*CytarabineLymphomatus meningitisFact-CNS
DynastatParecoxibPostoperative painPain Intensity Difference questionnaire
EmadineEmedastineSeasonal allergic conjunctivitisA nine-point symptom assessment patient diary scale
EmendAprepitantPrevention of nausea and vomiting associated with cancer chemotherapyFunctional Living Index—Emesis
ExelonRivastigmineAlzheimer's dementiaProgressive Deterioration Scale
EnbrelEtanerceptRheumatoid arthritisSF-36 Health Survey, Health Assessment Questionnaire
EVRANorelgestromin and ethinyl estradiolFemale contraceptionUnspecified scale to measure patient satisfaction
Fabrazyme*Agalsidase betaFabry diseaseSF-36 Health Survey, McGill Pain Questionnaire
FoscanTemoporfinCancerWashington University Head and Neck Questionnaire
FuzeonEnfuvirtideHIV-1Subcutanous Injection Survey
HerceptinTrastuzumabCancerEuropean Organization for Research on Treatment of Cancer QLQ (EORTC)
HumalogInsulin lisproDiabetesUnspecified HRQL and patient satisfaction measure
Humira*AdalimumabRheumatoid arthritisHealth Assessment Questionnaire (HAQ), Functional Assessment of Chronic Illness Therapy—Fatigue Scale (FACIT-F), Multidimensional Assessment of Fatigue (MAF), SF-36 Health Survey, EQ-5D, Health Utility Index
HycamtinTopotecanCancerEuropean Organization for Research on Treatment of Cancer QLQ (EORTC)
InsulatardInsulin human rDNADiabetesUnspecified HRQL and patient satisfaction measure
Intron AInterferon alfa-2bHepatitis, CancerUnspecified HRQL instrument
Ixense*Apomorphine hydrochlorideErectile dysfunctionInternational Index of Erectile Dysfunction, Brief Sexual Function Inventory
KeppraLevetiracetamEpilepsyUnspecified HRQL instrument
KineretAnakinraRheumatoid arthritisHealth Assessment Questionnaire
Kudeq*ValdecoxibOstheoarthritis, rheumatoid arthritis, primary dysmenorrheaWestern Ontario and McMaster Universities (WOMAC) OA Pain Index and OA Physical Function Index, Patient's Global Assessment of Arthritis, Patient's Overall Assessment of Pain measured by Visual Analog Scale (VAS), Patient's Global Assessment of Disease Activity, Modified Health Assessment Questionnaire Disability Index (mHAQ)
LantusInsulin glargineDiabetesDiabetes Treatment Satisfaction Questionnaire and Well-Being Questionnaire
Levitra*VardenafilErectile dysfunctionInternational Index of Erectile Function (IIEF), question 2 and 3 of the Sexual Encounter Profile, a single-item self-reported global assessment question, Fugl-Meyer Quality of Life Questionnaire, Center for Epidemiological Studies Depression Scale (CES-D)
LiprologInsulin lisproDiabetesUnspecified HRQL and patient preference measurement
LumiganBimatoprostChronic open-angle glaucoma and ocular hypertensionUnspecified patient satisfaction questionnaire
MyocetDoxorubicinCancerEuropean Organization for Research on Treatment of Cancer QLQ (EORTC)
Neoclarityn*DesloratadineRhinitisSF-36 Health Survey, rhinoconjunctivitis QoL questionnaire
NeorecormonEpoetin betaRenal failureUnspecified HRQL instrument
NeuroblocBotulinum toxin type BCervical dystoniaPatient's Global Assessment, self-assessed pain measures
NorvirRitonavirHIVThree different unspecified HRQL instruments and Karnofsky Performance Scale
NutropinAQSomatropinGrowth failureUnspecified HRQL instrument
OnsenalCelecoxibFamilial adenomatous polyposisSF-36 Health Survey
OpatanolOlopatadineSeasonal allergic conjuctivitisUnspecified HRQL instrument
OptisulinInsulin glargineDiabetesDiabetes Treatment Satisfaction Questionnaire and Well-Being Questionnaire
Opulis*DesloratadineRhinitisSF-36 Health Survey, Rhinoconjunctivitis QoL Questionnaire
PanretinAlitretioninKaposi's sarcomaUnspecified HRQL instrument
Patrex*SildenafilErectile dysfunctionInternational Index of Erectile Function and unspecified HRQL questionnaire
PaxanePaclitaxelAIDS-related Kaposi's sarcomaSymptom Distress Scale, Karnofsky Performance Score
PegasysPeginterferon alfa-2aChronic hepatitis CSF-36 Health Survey, Fatigue Severity Scale
PegIntron*Peginterferon alfa-2bHepatitis CSF-36 Health Survey
ProthapaneInsulin human rDNADiabetes mellitusUnspecified HRQL and overall treatment satisfaction measure
ProtopicTacrolimusAtopic dermatitisDermatology Life Quality Index
ProtopyTacrolimusAtopic dermatitisDermatology Life Quality Index
RayzonParecoxibPostoperative painPain Intensity Difference questionnaire
RebetolRibavirinHepatitis CUnspecified HRQL measure
RebifInterferon beta-1aHIV-1General Health Questionnaire, Epidemiologic Depression Mood Scale (CES-D), Beck Hopelessness Scale
RemicadeInfliximabCrohn's diseaseInflammatory Bowel Disease Questionnaire, IBDQ, Health Assessment Questionnaire, HAQ
SomavertPegvisomantAcromegalyUnspecified HRQL measure
StalevoLevodopa/ Carbidopa/ EntacaponeParkinson's diseasePatient preference for treatment
StocrinEfavirenzHIVMOS-HIV Health Survey, Karnofsy scale
SustivaEfavirenzHIVMOS-HIV Health Survey, Karnofsy scale
Taluvian*Apomorphine hydrochlorideErectile dysfunctionInternational Index of Erectile Dysfunction, Brief Sexual Function Inventory
Taxotere*DocetaxelCancerEQ-5D, Karnofsky performance status, Lung Cancer Symptom Scale, European Organization for Research on Treatment of Cancer QLQ (EORTC)
Temodal*TemozolomideCancerUnspecified HRQL measure
Thyrogen*Thyrotropin alfaCancerSF-36 Health Survey, Profile Mood States
Tikosyn*DofetilideAtrial fibrillationUnspecified HRQL instrument
Trudexa*AdalimumabRheumatoid arthritisHealth Assessment Questionnaire (HAQ), Functional Assessment of Chronic Illness Therapy—Fatigue Scale (FACIT-F), Multidimensional Assessment of Fatigue (MAF), SF-36 Health Survey, EQ-5D, Health Utility Index
Uprima*Apomorphine hydrochlorideErectile dysfunctionInternational Index of Erectile Dysfunction, Brief Sexual Function Inventory
ValdynValdecoxibOstheoarthritis, rheumatoid arthritis, primary dysmenorrheaWestern Ontario and McMaster Universities (WOMAC) OA Pain Index and OA Physical Function Index, Patient's Global Assessment of Arthritis, Patient's Overall Assessment of Pain measured by Visual Analog Scale (VAS), Patient's Global Assessment of Disease Activity, Modified Health Assessment Questionnaire Disability Index (mHAQ)
Vaniqa*EflornithineDermatologySubjects Self-Assessment Questionnaire (an instrument with six questions each rated by a 10 cm visual analog scale)
VentavisIloprostPrimary pulmonary hypertensionUnspecified HRQL instrument
Viagra*SildenafilErectile dysfunctionInternational Index of Erectile Function, Global Assessment Question, Partner Questionnaire, and an unspecified HRQL questionnaire
ViraceptNelfinavirHIVKarnofsky Performance Scale
ViraferonInterferon alfa-2bHepatitis B and CUnspecified HRQL instrument
ViraferonPegPeginterferon alfa-2bHepatitis CSF-36 Health Survey
Vivanza*VardenafilErectile dysfunctionInternational Index of Erectile Function (IIEF), question 2 and 3 of the Sexual Encounter Profile, a single-item self-reported global assessment question, Fugl-Meyer Quality of Life Questionnaire, Center for Epidemiological Studies Depression Scale (CES-D)
XelodaCapecitabineCancerEuropean Organization for Research on Treatment of Cancer QLQ (EORTC)
XenicalOrlistatObesityPatient rated distress, depression, self-regard, and satisfaction with treatment
ZavescaMiglustatType 1 Gaucher diseaseUnspecified HRQL instrument
ZyprexaOlanzapineSchizophreniaQuality of Life Scale
Zyprexa VelotabOlanzapineSchizophreniaQuality of Life Scale

In the majority of cases (i.e., 55), the PROs used to evaluate these products were various kinds of HRQL outcomes. In 46 cases, PROs other than (or in addition to) HRQL outcomes were reported. PROs other than HRQL included, patient-reported symptoms, functional status, pain, disability, bother, patient preference, patient satisfaction with treatment, and patient's global assessment.

A subgroup analysis was performed on the Anatomical-Therapeutic-Clinical (ATC) groups reflecting those 67 products that had different active ingredients or had the same active ingredients but had different therapeutic indications, forms, or manufacturer (14 out of the 81 products were identical but were registered under different trademark names). As Figure 1 reveals, the most prominent disease areas where HRQL and other PROs were included in the drug evaluation were antineoplastic and immunomodulating agents, mainly involving cancer products.

Figure 1.

The relative share of products in different ATC drug groups of all products with HRQL and other PRO data in regulatory documents.

The Scientific Discussions varied in the level of detail provided on the HRQL and other PRO outcomes. Specification of these variables as primary or secondary trial endpoints was not always clearly stated in the Scientific Discussion document. Based on the information available, these outcomes were generally included as secondary efficacy endpoints. PROs appeared as primary endpoints mainly in those cases where they were the only possible primary efficacy parameters, for example, self-assessed pain or the effectiveness of treatment of erectile dysfunction.

In 24 of the 81 Scientific Discussions (30%), the specific HRQL or other PRO instrument used was not mentioned. The Scientific Discussion simply referred to the results, for example, as “quality of life” or “patient satisfaction” data. There was a tendency not to specify the instrument if HRQL results did not differ significantly between the treatment groups. The majority of the reported clinical trials used condition-specific instruments to measure HRQL outcomes. The only generic instrument used more than once was the SF-36 Health Survey. The SF-36 was used in the evaluation of drugs for the treatment of rhinitis, rheumatoid arthritis, popyposis, Fabry disease, cancer and chronic hepatitis C. The SF-36 was most often used along with a disease-specific HRQL or other PRO instrument. In the case of rhinitis and cancer, results using condition-specific instruments paralleled results using the SF-36. In the case of Fabry disease, the drug Fabrazyme did not show statistically significant improvements on the SF-36 scores but significant results were detected on McGill Pain Questionnaire. Preference-based measures, the EQ-5D and the Health Utility Index, were reportedly used for Humira and Taxotere, but the documents did not include sufficient information to judge the role of these data in evaluating the effectiveness of the product. The appearance of HRQL and other PRO data in the Scientific Discussion of products indicated for the treatment of cancer, osteoarthritis, rheumatoid arthritis, Parkinson's disease, Crohn's disease, and obesity were consistent with EMEA/CHMP recommendations. The remaining products were reviewed and approved with HRQL and other PRO data in the absence of any specific disease-group level recommendations. These results were consistent with interview data suggesting that the EMEA will consider HRQL (or other type of PRO) data submitted in support of drugs where the relevant EMEA/CHMP recommendation does not explicitly mention these endpoints.

Summary of product characteristics.   Of the 81 products with HRQL or other PRO data in the Scientific Discussion, 26 (32%) included these results in the SPC document. With the exception of Aldurazyme, Depocyte and Taxotere, SPCs included HRQL or other PRO data only if the results were statistically significant. In the case of Taxotere, equivalent HRQL outcomes across treatment groups were interpreted as a positive outcome, an indication that treatment did not adversely affect patient HRQL. In general, HRQL data were presented very briefly, without specific results or domain level information. The only exception was the SPC of Humira (also registered as Trudexa) that contained more detailed information on PRO instruments used and results observed on different subscales of these instruments:

Health-related quality of life and physical function was assessed using the disability index of the Health Assessment Questionnaire (HAQ) in all four adequate and well-controlled trials, which was a prespecified primary endpoint at week 52 in Study III. All doses/schedules of Trudexa in all four studies showed statistically significantly greater improvement in the disability index of the HAQ from baseline to Month 6 compared to placebo and in Study III the same was seen at Week 52. Results from the Short Form Health Survey (SF 36) for all doses/schedules of Trudexa in all four studies support these findings, with statistically significant physical component summary (PCS) scores, as well as statistically significant pain and vitality domain scores for the 40 mg every other week dose. A statistically significant decrease in fatigue as measured by functional assessment of chronic illness therapy (FACIT) scores was seen in all three studies in which it was assessed (Studies I, II, IV).

Package leaflet.  Only products containing the active substance desloratadine included information on HRQL or other PRO effects of treatment in the package leaflets. This could be due to the fact that package leaflets approved by the EMEA focus on mainly on safety issues, usage, side effects, and storage information, rather than efficacy data.

As the following paragraph illustrates, the package leaflet for Aerius (desloratadine) included information from the findings of the rhinoconjunctivitis-specific HRQL measurement (activities, sleep) used in the clinical trials:

What Aerius is and what it is used for

. . . Aerius is an antiallergy medicine that does not make you drowsy. It helps control your allergic reaction and its symptoms. Aerius relieves symptoms associated with seasonal allergic rhinitis (for example, hay fever) such as sneezing, runny or itchy nose, itchy palate, and itchy, red or tearing eyes. Aerius is also used to relieve the symptoms associated with chronic idiopathic urticaria such as itching and hives.

Relief of these symptoms lasts a full day and helps you to resume your normal daily activities and sleep.

Abstracts.  Eight products included reference to HRQL and other PRO data in their abstracts. Five of these contained the active substance of desloratadine and were indicated for the treatment of rhinitis. The remaining three products were Caelyx indicated for the treatment of AIDS-related Kaposi's sarcoma, Depocyte for the treatment of lymphomatus meningitis, and Thyrogen used in thyroid cancer patients.

The abstract for Depocyte included a statement on potential quality of life improvement (due to the ease of administration) that was not supported by the HRQL evidence introduced in the Scientific Discussion. The abstract stated: “Depocyte has a more convenient schedule of administration compared to conventional ara-C as it reduces the need for multiple injections and this may impact quality of life favorably.” The Scientific Discussion and the SPC indicated HRQL outcomes were compared between the two groups and no differences were found. It was unclear, however, if the term “quality of life” was used in the same meaning across the SPC and the abstract.

Products containing desloratadine appeared to be the only case in which the Scientific Discussion, the SPC, the package leaflet, and the abstract all included some of the HRQL and other PRO results.

Numeric Summary

The development of disease-specific instruments and new research and statistical methods for HRQL and PRO evaluation is a rapidly developing research area. Therefore, a relevant question is whether there is any indication of an increasing role that these health outcomes play in the drug regulatory process. Temporal trends involving recommendations on HRQL and PRO data in EMEA/CHMP guidance documents is as yet difficult to judge because these documents have been developed and repeatedly revised over the past 9 years of the operation of the EMEA. Regarding the authorization of pharmaceutical products, Figure 2 illustrates that with the increase of registered products, the number of products with HRQL and other PRO data in their regulatory documents also increased and exceeded 30% in the last 4 years (between 1999 and 2003).

Figure 2.

The number of registered different drugs and number of HRQL or other PRO data in regulatory documents. Note: “Different drug” means those drugs that have different active agent or those that include the same active agent but have different therapeutic indication. Multiple registered drugs with identical active agent and identical therapeutic indication were counted only once.

Discussion and Conclusions

The review revealed that although no general regulatory guidance documents have been issued for HRQL and other PRO evaluation by the EMEA, more than half of the EMEA/CHMP guidance documents on clinical investigation of pharmaceutical products in specific disease areas include reference to, or recommendations on, these outcomes. Even in the absence of general or disease-specific guidance, HRQL and other PRO data appeared in the Scientific Discussions of 34% of the products submitted between 1995 and 2003. Claims appeared in the SPCs of some of the products but relatively little detail or information is included in the SPC. These descriptive findings should be carefully interpreted within the light of the evolving European regulatory environment and current measurement and methodological developments within the HRQL research field.

During the past decade, independent health outcomes researchers in Europe and North America have outlined recommendations on HRQL evaluation for drug regulatory purposes [5,39–42]. Although these recommendations have played an important role in setting standards for good science in HRQL research, they did not constitute the official opinion of any regulatory authorities. The direct adaptation of these scientific recommendations by a regulatory body like the EMEA would not have been easily feasible, not least because of the variations reflected across them in some conceptual and methodological issues such as the definition of HRQL. As the interviews with the EMEA representatives confirmed, it would have been difficult for a regulatory authority to develop prescriptive guidelines for an outcomes research community which is characterized by ongoing debates on important research issues.

Historically, the rapidly evolving HRQL research environment have been an contributing factor to the lack of an official harmonized approach for considering HRQL outcomes in the newly created centralized European regulatory review process. Because of a lack of a specified EMEA guidance documents, decisions on considering HRQL data during authorizations were delegated to EMEA appointed committees consisting of external experts with varying degrees of experience and expertise in HRQL or other PRO research. The potential inclusion of HRQL and other PRO data in official guidance documents on regulatory clinical trials for each disease area has been dependent on individual experts’ views and knowledge about HRQL and PRO research who have been involved in the drafting and approval of guidance documents. This has lead to variations and inconsistencies in recommendations, in terms of positioning HRQL and other PRO data among study endpoints, requirements on psychometric standards, analytical requirements, and other aspects of HRQL evaluation.

On the other hand, our review of performed authorizations revealed that, despite these uncertainties in published recommendations on HRQL and other PRO data, manufacturers opted to collect and submit HRQL and PRO data in their clinical trials and regulatory filings. The appearance of these data in the approved regulatory documents of these products reflected the EMEA's approach that all HRQL and PRO data can be considered, given that the data inclusion is justified and the scientific methods used were rigorous. The continued application of HRQL and PROs in industry sponsored research is driven by the need to demonstrate differentiation among competing products and in demonstrating value in terms of patient health outcomes relative to expenditures.

An important area for the more efficient use and communication of HRQL and other PRO data would be to better define the optimal role of HRQL and PRO data in the published regulatory documents of pharmaceutical products. The purpose of the Scientific Discussion is to reflect how authorization decisions were made and what evidence they were based on. Nevertheless, currently it is not transparent how submitted HRQL and PRO information is screened, considered, and abstracted into the Scientific Discussion document. Because of the lack of publicly available information on the screening process, it is unclear from the review of documentation from the EMEA/CHMP how judgments are made regarding the quality of PRO measures that are used in clinical trials. It would seem that results from different measures are taken mainly at face value without thorough review of the validity and reliability of different measures. It is also not clear how judgments are made about the translation and cultural adaptation process that is required to use PROs in different countries.

The objective of the SPC and the Package Leaflet is to inform physicians and patients about the medication. According to the current regulations, safety information has priority over efficacy data and the avoidance of any potential promotional use of information [43]. No efficacy data is generally expected in the SPC and no efficacy data is allowed in the Package Leaflet apart from exceptional cases. Our review revealed no clear tendency for HRQL and other PRO data that are regarded as sufficiently important to be included in the SPC and the Package Leaflet.

Finally, the results of this review suggest that there is some evidence for the varying use of HRQL and other PRO information in the regulatory process across disease groups. There are areas, cancer in particular, where official recommendations suggest a potentially important role for HRQL data and a high proportion of the submitted regulatory applications in oncology include these PROs. On the other hand, in several disease areas with well-established HRQL research and instruments, official recommendations lack reference to HRQL and other PRO data and regulatory documents also suggest the lack of consideration of this important aspect of the treatment in the regulatory process.

Health outcomes researchers view HRQL and other PRO data as useful in understanding the impact of disease and treatment on patient outcomes. These outcomes data can assist physicians and their patients in safe and effective treatments. The application of these health outcome measures should not be dependent on uncertain factors, such as lack of understanding between health outcomes researchers and regulatory decision-makers. Health outcomes researchers should provide evidence for including HRQL data in clinical trials, highlight the added value of HRQL results to the clinical data for the regulatory decision-makers, and should, in general, follow good scientific methods for the design, conduct and analysis of clinical trial data. The regulatory decision-makers could better communicate their evaluation standards and process related to HRQL data and develops and updates harmonized procedures and expertise to adequately appraise and use the HRQL and other PRO information in the regulatory decision-making process. The just recently released reflection paper on HRQL research by the EMEA and its currently ongoing public debate is a promising step towards achieving this aim [44].

The authors would like to thank Andrew Lloyd, Wendy Schneider, Jack Boulton, and Steven Kimberley for their useful help in the preparation of this article. The authors also thank Dr. Olivier Chassany on behalf of ERIQA, Dr. Richard J. Willke, and two anonym referees for valuable comments, from which the article benefited considerably.

Source of financial support: This research was supported in part by the following companies: Novartis, Sanofi, Merck, Pfizer, NovoNordisk, Roche, 3M, Wyeth, Pharmacia, Schering Plough, and Eli Lilly.

Ancillary