Chronic fatigue syndrome (CFS) is a debilitating condition that affects over 1 million adults in the United States with many more experiencing chronic symptoms of fatigue [1-4]. A CFS diagnosis is based on self-reported chronic symptoms, unexplained severe fatigue present for six months or more, and at least four of eight other symptoms (impaired memory or concentration, muscle pain, pain in multiple joints, unrefreshing sleep, tender lymph nodes, sore throat, headaches, post-exertional malaise) that are severe enough to significantly interfere with the performance of daily activities and work [5]. Diagnosis of CFS requires ruling out concurrent medical and psychiatric conditions with similar clinical signs and symptoms. CFS has no diagnostic biomarkers, signs, or laboratory tests, and despite over two decades of research, the etiology and pathophysiology are yet to be determined.

The lack of objective diagnostic criteria poses a challenge for physicians and helps explain the low diagnostic confidence [6]. In spite of this, most physicians (94%) are aware of CFS, and 40% report having diagnosed the illness [7]. This high level of provider awareness belies estimates that fewer than 20% of people with CFS have been diagnosed [8].

CFS patient registries can provide information concerning the nature of the illness among those receiving medical care. Twin registries and the Gulf War Registry have been used to study CFS prevalence, the medical impact of illness, and the role of environmental and genetic factors [9-13]. However, published twin registries have not medically evaluated participants. They use questionnaires to assess and classify fatigue and CFS often in conjunction with reported physician diagnosis [12,13]. Registries, such as the United Kingdom pilot CFS/ME registry, also provide a cost effective means of evaluating aspects of CFS, identifying CFS for clinical studies, and evaluating clinical course [14]. In conjunction with population-based CFS surveillance, registries provide a means of exploring factors associated with health care use by people with CFS.

This report concerns the feasibility of establishing and operating a pilot registry that was conducted in Bibb County, Georgia. The objectives of this registry were 1) to recruit and register medical and ancillary healthcare providers in and around Bibb County, 2) to encourage enrolled healthcare providers to refer patients with unexplained fatiguing illnesses and CFS, 3) to clinically evaluate referred patients, and 4) to implement an education outreach initiative. The registry was later supplemented by local support group members through self-referral. We clinically evaluated all referred patients according to the protocol used in population surveillance so as to diagnose CFS by international criteria [5], identify exclusionary medical and psychiatric conditions, and collect sociodemographic and related information.

The pilot registry was conducted in Bibb County, Georgia from September 2008 to March 2010. This pilot included a six-month healthcare provider recruitment phase followed by a full year of patient referrals to a research clinic, where patients were clinically evaluated [1]. A needs assessment in the form of focus groups with Bibb County healthcare providers was conducted before the registry, and these results helped to inform the registry study design. For example, time commitments and burden on support staff were identified as concerns, whereas a benefit was the desire to receive updates throughout the study period. In December 2009 the Registry expanded to allow members of the local CFS support group to self-refer to the registry. This study adhered to U.S. Department of Health and Human Services policies for the protection of human research subjects and was approved by the CDC Institutional Review Board. All participants gave informed consent. The registry methods are reported in five parts: Healthcare Provider Recruitment, Healthcare Provider Enrollment and Referrals, Support Group Referral, Referrals for Clinical Evaluation, and Registry Education Initiatives.

The pilot registry exceeded the recruitment goal of 600 by registering 827 healthcare providers. Despite multiple mailings, CME events and a registry website, telephone recruitment was the most successful method for provider registration, accounting for 94% of provider registration. One reason that telephone recruitment worked well is that healthcare providers could immediately respond to verification questions upon receipt of the phone call. Focus groups had suggested that a registry website would be desirable for provider registration; nevertheless, only 2 of 827 providers registered through the site. However, enrolling through the website, fax, or postal mail required a two-step process: 1) filling out the Provider Verification Form and 2) submitting the information. Given time constraints and other practice priorities, providers may have been dissuaded to register through these channels.

Despite success in recruiting healthcare providers, only 42 (5%) of the 827 enrolled providers referred patients for clinical evaluation, even though referral criteria were relaxed and based on fatiguing illnesses and not just CFS (Table 1). Several explanations may account for this outcome. First, healthcare providers were asked to refer patients that came into their offices during the study enrollment period and not to look in medical charts for referrals, thereby missing patients who met inclusion criteria. Second, we opened registry registration to healthcare providers of all specialties and not just to those who see or treat people with fatigue. Many specialty providers may not have had the opportunity for referral, as their patient population would not be seeking care for fatigue. Supporting this concept is the fact that none of the enrolled healthcare providers from anesthesiology, allergy and immunology, infectious disease, rheumatology, endocrinology, gastroenterology, neurology, surgery, orthopedics, or ophthalmology made referrals, whereas primary care providers (family practice and internal medicine) generated the greatest number of referrals (n = 33, 38%). This finding also supports the view that many people with unexplained fatiguing illnesses continue to seek healthcare from their primary care providers. Finally, the firm hired to recruit providers was compensated per enrolled provider and therefore had an incentive to recruit and enroll as many providers as possible, including providers who may not have had an interest in chronic fatigue or the registry.

Of the 42 providers who made referrals, 62% received an outreach visit or attended either a focus group, CME, or “Meet and Greet” event. Of healthcare providers who received education or outreach, three-quarters referred patients and two-thirds had patients complete the clinical evaluation. Thus, education and active outreach to providers may have contributed to a greater number of patient referrals and clinic evaluations as compared to no participation or providers who did not receive these initiatives. When comparing education events (focus groups, CME, or “Meet and Greet”) to outreach activities (visits or visits with incentives), outreach resulted in a higher number of referrals (49%) and clinic evaluations (44%) compared to education. Researchers interested in developing and operating a registry in which referrals are a goal may want to consider incorporating visits to healthcare providers as one technique to increase patient referral rates.

We opened registry enrollment to local support group members. Although the self-referral process worked well, with no complaints or confusion, only 28% of the support group referred themselves for screening. Despite the need for further research, results from this pilot registry suggest that future CFS registries may want to consider including enrollment directly from patient populations if full clinical and laboratory exams are included in a clinical assessment.

One-hundred and four patients were clinically evaluated and, of these patients, 37 had CFS, 18 ISF, and 49 an exclusionary condition. A little more than half of the sample reported a previous CFS diagnosis, but this history was equally distributed across illness classification conditions. A relatively high proportion (47%) of the registry patients had exclusionary conditions identified after clinical evaluation. This finding replicates other CFS studies in which 50 - 70% of patients who met symptom criteria were found to have medical or psychiatric exclusions [3,16,17]. Future research needs to examine the potential for detecting treatable medical exclusionary conditions among people suffering from fatiguing illnesses because treatment may improve fatigue and associated symptoms for many of these patients.

More revealing was the proportion (59%) of exclusions found in the support group as compared to the provider referred group. In fact, 63% of the support group patients evaluated (29/46) had an exclusionary condition identified. Despite awareness of their illness (as evidenced by membership in support group), the self-referral group had undetected medical conditions, and some patients experienced more than one exclusionary condition. For example, this group had seven cases of diabetes compared to one in the provider referral group, and seven cases of hypothyroidism compared to three in the provider group. In both groups, many of the exclusionary conditions were detected on the screening laboratory tests and are treatable.

Forty-one percent of provider referrals were determined to have an exclusionary condition. Because providers were given a medical exclusion list and knew the medical history of patients, one would expect lower exclusion rates, e.g., few provider referrals resulting in exclusions. Two possible explanations address this phenomenon. First, most of the exclusions were identified on the basis of laboratory tests. Physician providers were asked to give potential referrals by running basic lab tests and completing a medical history; however, referrals could be made without the requirement for a complete work-up, and laboratory testing may not have been conducted. Lack of basic lab tests may have been especially true for non-physician referrals, for which the requirement was a medical history. Second, providers may have welcomed participation in the registry as a method to get assistance for diagnosing difficult or unknown etiologies. Finally, the other main category of exclusions for provider referrals were bipolar disorder, major depressive disorder, substance abuse, and anorexia bulimia, which is understandable as some of these diagnoses are difficult to establish.

Sixty-five percent of provider-referred patients had CFS compared to 35% of support group referrals. Sixty-five percent of referred patients had CFS, which replicates similar findings from the CFS Rochester Epidemiology Project in which 66% of patients were accurately coded into the CFS category [17]. For an illness that has no known etiology or biomarkers, a fairly high level of knowledge and awareness was shown for providers who referred two-thirds of patients diagnosed with CFS at clinical evaluation. Although further research is warranted, part of the awareness may be due to the education and outreach events in which 62% providers participated. The main reason for not meeting CFS criteria was exclusionary conditions as opposed to being classified as ISF. A higher proportion of provider-referred patients (78%) were classified as ISF as compared to support group patients (22%). After exclusionary conditions are fully treated, those patients otherwise meeting criteria for CFS would be managed in the same way as CFS.

We aimed to increase awareness and knowledge of CFS among the healthcare community through several education initiatives. CME events offered the opportunity for professional credits and were offered through local partnerships. We found from anecdotal feedback that the newsletter, an item identified through the needs assessment, was well-received and appreciated by providers. As community-based research has shown [18], one of the most valuable aspects of this registry, which contributed to its success, was the partnership with the local medical society. The partnership allowed for educational outreach events with CME that would otherwise have been time consuming and difficult to implement. The partnership also gave registry staff access to the healthcare community and enabled us to participate with the community in building the registry.

This research has several limitations. The results reflect the experience of one study and future fatigue and CFS registries may have different recommendations, as registries have different study designs. The total number of patients evaluated (104) is relatively small and the pilot was conducted in one locality, so the findings may not be generalizable. We did not have a way to determine the number of potentially eligible patients that each provider saw during the time of the study. Because this study was a pilot registry, we did not determine if longitudinal information could be obtained.

While we did not exclude severely ill patients from study participation, the study design made it difficult for home-bound or other severely ill patients to participate as they would need to travel to the local clinic for the 8-hour testing day. In fact, two referred patients canceled their clinical evaluations because they were not feeling well enough to participate (Table 3). Researchers may want to consider this aspect of participant recruitment for future registries. Extra costs or other methodological issues may be incurred in arranging for testing at a patient’s house, but this would allow for home-bound and severely ill patients to be included in the registry.

Finally, depending on recruitment methods, cost and sustainability of registries are an important aspect of further research. For example, this study demonstrated that in the Bibb County, Georgia area telephone recruitment was the best method to enroll providers. We used several different methods, for example, mailing materials and a study website, before turning to telephone recruitment even though focus group results indicated recruitment by website was a preferred option. When developing future registries, researchers may want to consider various recruitment methods in order to sustain enrollment. All studies incur costs in the form of laboratory testing, medical staff time, and data collection; study objectives should be prioritized in terms of cost and sustainability.

In the needs assessment phase, providers had expressed concerns about the 1996 Health Insurance Portability and Accountability Act (HIPAA) procedures and the preference for not being involved in the process of consenting patients. These concerns proved challenging in implementing a system in which providers screened potential patients, but the patients had to complete the referral process. From the provider referrals, the registry staff could not contact 11% of patients for clinical assessment because 10 people did not follow through by submitting permission-to-be-contacted forms. This finding may also serve as a challenge to other registries that rely on physician referrals or access to medical records to confirm a diagnosis.

A pilot registry of unexplained fatigue and CFS in which healthcare providers were registered to refer patients suspected of having unexplained fatiguing illness or CFS for a standardized clinical evaluation and diagnosis, provided information on clinicians’ accuracy in recognizing aspects of fatiguing illness and CFS. It also identified gaps in their knowledge, which may inform educational initiatives and ultimately increase the number of people receiving treatment. This pilot registry design allowed many different types of healthcare providers to participate and permitted tracking of medical specialty in terms of the referral process. The registry demonstrated the feasibility of including both provider and support group referrals. Results highlight the need and importance for inclusion of thorough clinical and laboratory exams in a CFS registry rather than relying on report of a previous CFS diagnosis. Many exclusion conditions were detected by laboratory tests and appeared to be unknown to the participating patients and healthcare providers. The number of untreated conditions in this population also emphasizes the need for continued medical follow-up to identify and manage conditions that may develop, rather than attributing all illness symptoms to CFS. Future CFS registries should consider how to address exclusions when enrolling potential patients.

CFS: Chronic fatigue syndrome; CDC: Centers for Disease Control and Prevention; ISF: Insufficient fatigue or symptoms; VA: Veteran’s Affairs; CME: Continuing medical education; CATI: Computer-assisted telephone interview; CAPI: Computer-assisted personal interview; BMI: Body mass index; SCID: Structured clinical interview for DSM disorders; HIPAA: Health Insurance Portability and Accountability Act of 1996.

The authors declare that they have no competing interests.

DJB was Co-Principal Investigator and EM was Principal Investigator for the pilot registry. DJB and EM were responsible for designing the study, developing the study protocol, monitoring protocol implementation, and planning data analysis. RD, CN, LL, SR, and JK were involved in the study protocol implementation, day-to-day administration of study protocol, and data/specimen collection. JFJ and RB assisted in study design and development of study protocol, reviewed medical data, assessed eligibility criteria, and along with JSL, classified study subjects. JSL and ERU were involved in study design and review of the study protocol. JSL and HT were responsible for database development and administration. All authors were involved in the preparation of this paper and approved the final manuscript.