The prevalence and incidence of medical conditions in healthy pharmaceutical company employees who volunteer to participate in medical research

Abstract

Aims

Although clinical research in healthy volunteers is commonly performed there have been few studies of the value of the medical screening of subjects. The aim of this study was to investigate the prevalence and incidence of medical conditions found during the medical screening of ‘healthy’ subjects employed in a pharmaceutical company who volunteered to participate in medical research.

Methods

This was a retrospective study of the medical notes of all the subjects who volunteered for membership of the Zeneca Clinical Pharmacology Unit’s healthy volunteer panel over a 4 year period from 1990 to 1994. The prevalence of medical conditions found at presentation was determined. The incidence of medical conditions during the 4 year observation period was also ascertained. Medical screening included a full medical history and examination, clinical chemistry, haematology and urinalysis screens, pulmonary function tests, ECGs, 24 h ambulatory cardiac monitoring and a request for information from the volunteer’s General Practitioner.

Results

Prevalence-1293 subjects volunteered to join the panel of which 156 subjects (12%) were not accepted at presentation including 141 (10.9%) for medical reasons. The most medical common reasons were; previously diagnosed medical conditions (3.3%), cardiovascular abnormalities (1.9%), abnormal liver function tests (1.9%), anaemia (1.2%), hyperlipidaemia (1.1%), excess alcohol intake (0.6%) and thyroid disease (0.5%). Incidence—36 of the 1137 volunteers (0.8% per year) accepted onto the panel subsequently developed medical conditions of which the most common were; anaemia (0.29% per year), cardiovascular abnormalities (0.13% per year) and vasovagal syncope (0.13% per year).

Conclusions

This study demonstrates the importance of medical screening before healthy volunteers participate in clinical research.

Introduction

Clinical research in healthy volunteers is commonly undertaken, but there have been few studies of the value of medical screening of subjects. Such studies are important because serious and even life threatening events are known to occur during clinical research in healthy subjects [1]. Additionally volunteers have died during [2] or after [3] participation in research involving therapeutic agents.

Watson & Wyld found that medical screening (including electrocardiogram, clinical chemistry and haematology and urinalysis screens) of volunteers led to 28.5% of applications being rejected for all reasons including 14% on clinical grounds [4]. There has been no other study of the prevalence of medical conditions in healthy subjects who volunteer to participate in medical research. Additionally, many industrial and academic clinical pharmacology units operate ‘healthy volunteer panels’ including their own employees in which panel members participate regularly in clinical research. The value of health screening in this population has not been studied.

We have undertaken a retrospective epidemiological study to assess the prevalence of medical conditions in healthy subjects who volunteer to participate in medical research and the incidence of such conditions in a ‘healthy volunteer panel’.

Methods

This was a retrospective study of the medical notes of all the subjects who volunteered for membership of the Zeneca healthy volunteer panel over a 4 year period from November 1990 to November 1994. The period prevalence over 4 years of medical conditions found in 1293 subjects who volunteered to join the healthy volunteer panel was obtained (the period prevalence of a disease is defined as the proportion of a population that are cases at any time within a stated period [5]).

The incidence of medical conditions occurring over the 4 years observation period, in those subjects who were accepted on to the healthy volunteer panel, was determined (the incidence of a disease is defined as the rate at which new cases occur in a population during a specified period [5]). Volunteers were all employees of Zeneca Pharmaceuticals who had been recruited by advertising within the workplace. All were aged between 18 and 62 years.

Screening of subjects for membership of the healthy volunteer panel was carried out by the physicians of the Zeneca Clinical Pharmacology Unit. All volunteers are required to undergo a medical screening process for panel membership. Medical screening consists of the following stages:

1. medical history and examination

   cardiac investigation with a computerized 12 lead ECG and 2 lead 24 h ambulatory cardiac monitoring using a Schiller ECG recording and analysing machine

   lung function assessment with spirometry and flow volume loop.

   laboratory investigations—clinical biochemistry including electrolytes, liver function, lipids and thyroid function.

   —clinical haematology including full blood count and coagulation.

   —hepatitis B virology

   —urinalysis

   request for information from the subjects General Practitioner (GP).

The previously reported guidelines [6] for the management of 24 h ambulatory cardiac monitor (ACM) recording abnormalities in healthy volunteer studies were followed. The guidelines detail the appropriate management for healthy subjects with cardiac arrhythmias. These guidelines were derived from studies undertaken in our unit on ACM recordings which showed that only 13% of these in healthy subjects are entirely normal (i.e. sinus rhythm throughout). The report deals extensively with the type and prevalence of cardiac arrythmias on ACM recordings in healthy volunteers.

Laboratory values from 1000 subjects screened for membership of the healthy volunteer panel in the 4 years before November 1990 were used to establish reference ranges for the population who volunteer to participate in this Clinical Pharmacology Unit’s studies. Normal ranges for laboratory investigations usually cover 95% of the population [7]. Our reference ranges covered 98% of subjects, which avoids classifying 5% of previously healthy individuals as having laboratory abnormalities (i.e. we would anticipate 2% of our healthy volunteers being outside the reference range for any reason found on screening). An abnormal test result was repeated and volunteers were not excluded from the panel for isolated or spurious abnormalities in laboratory values.

GPs were informed of the reason for volunteers not being accepted onto the panel and where appropriate specialist referral was made for further investigation and/or treatment.

Results

Prevalence

One thousand two hundred and ninety-three subjects volunteered to join the healthy volunteer panel and of these a total of 156 subjects (12%) were not accepted; 141 (10.9%) were rejected for medical reasons and 15 (1.1%) were rejected for non medical reasons. Table 1 lists the medical reasons. Previously diagnosed medical conditions, cardiovascular abnormalities, abnormal liver function tests, anaemia, hyperlipidaemia, excess alcohol intake and thyroid disease all occurred with a prevalence greater than 0.5% (1 in 200 subjects). Of the 16 subjects found to be anaemic 13 were female and 3 were male.

Table 1.

Prevalence (1990–1994). Medical conditions at presentation in 141 volunteers rejected for panel membership.

Details of the previously undiagnosed medical conditions found at medical screening (that is all the above groups excluding previously diagnosed conditions) are given in Table 2. These 99 volunteers’ ages are presented in Figure 1a which shows the percentage of subjects with medical conditions stratified by age. Figure 1b shows the percentage of the total volunteers stratified by age. These distributions have different modal values; 41–45 years for Figure 1a and 31–35 years for Figure 1b. Additionally, the percentage of volunteers in each age band after 35 years is higher in Figure 1a than Figure 1b. This indicates that the prevalence of previously undiagnosed medical conditions is greater in the older age groups.

Table 2.

Previously undiagnosed medical conditions in 99 volunteers not accepted onto panel.

Figure 1.

a) Age stratification of 99 subjects with previously undiagnosed medical conditions and b) age stratification of total number of subjects (1293) who volunteered to join the healthy volunteer panel.

The most common non medical reason for rejection was difficulty in obtaining blood samples by venepuncture at medical screening, since this would present difficulties in obtaining blood samples for research purposes.

Incidence

During the 4 year observation period 36 of the 1137 volunteers (0.8% per year) accepted onto the panel subsequently developed medical conditions which excluded them from panel membership. Table 3 lists the conditions found. Anaemia, cardiovascular abnormalities and vasovagal syncope occurred with an incidence of greater than 0.1% per year.

Table 3.

Previously undiagnosed medical conditions in 36 volunteers previously accepted onto the panel.

Six volunteers (0.13% per year) developed vasovagal syncope at venepuncture during clinical studies—this occurred despite specific questioning at medical screening about fear of medical procedures and undergoing venepuncture for blood samples. One of these volunteers had an episode of loss of consciousness for approximately 2 min. He was hypotensive and his ECG showed a sinus bradycardia. He recovered without any medical intervention. In the other five volunteers there were less severe episodes with transient weakness and light headedness.

The volunteer who developed multiple sclerosis was diagnosed initially by her General Practitioner. Otherwise all the other medical conditions which led to the exclusion of volunteers from the panel were diagnosed as a result of the medical screening procedures. All 13 volunteers who became anaemic had donated blood for medical research. Ten of these volunteers were female and three were male.

Discussion

In our sample of 1293 subjects who volunteered to participate in medical research the prevalence of medical conditions was 10.9%. During the 4 year observation period the incidence of medical conditions developing in the 1137 subjects accepted onto the healthy volunteer panel was 0.8% per year.

These results indicate that medical screening for healthy volunteers before they participate as subjects in clinical research is important, both on health and safety grounds to ensure that subjects with undiagnosed medical conditions are not put at risk, and also on scientific grounds to ensure that the subject population studied is indeed composed of ‘healthy volunteers’. The commonest undiagnosed medical conditions (with a prevalence of greater than 0.5%—1 in 200) found in this study were cardiovascular disease, liver disease, anaemia, hyperlipidaemia, excess alcohol consumption and thyroid dysfunction. These conditions accounted for approximately 80% of the abnormalities detected.

Screening for renal disease by urinalysis and blood urea, creatinine and electrolyte levels forms part of almost all healthy volunteer screening procedures involving studies of experimental drugs. In our sample only two subjects (0.15%) were found to have haematuria on urinalysis and only one subject (0.08%) was found to have an abnormal blood test (raised creatinine). Thus the prevalence of renal conditions was 0.23%. Similarly, a full blood count is almost always performed before studies involving new drugs. In our series although 1.2% of subjects were found to have anaemia only two subjects (0.15%) had other haematological abnormalities (one subject with pancytopenia and one subject with eosinophilia). Respiratory function testing in which flow volume loops were obtained found one subject with undiagnosed chronic obstructive airways disease. Thus the prevalence of respiratory disease diagnosed by respiratory function testing was the same (one subject-0.08%) as that found by screening for renal disease by blood urea, creatinine and electrolytes and similar (0.08% compared with 0.15%) to the prevalence of renal disease detected by urinalysis and abnormalities of white blood cells detected by cell counts. This is a surprising result as all medical screening of subjects before they take experimental drugs will include urine and blood tests of renal function and white blood cell counts but few (if any) will include respiratory function testing.

Medical screening discovered previously undiagnosed medical conditions in subjects at all ages from 18 to 62 years. The age stratification of these volunteers with previously undiagnosed medical conditions is different to that of the total number of volunteers; there is a higher proportion of the former group above 35 years of age compared with the proportion of the total volunteer population above 35 years. This indicates that the prevalence of previously undiagnosed medical conditions in subjects who volunteer to participate in clinical studies increases at above 35 years of age.

Cardiovascular abnormalities were the most prevalent undiagnosed medical condition found at medical screening. Hypertension was the commonest condition. Hypertension [10], hyperlipidaemia [11, 12] and hypothyroidism [13] are all associated with increased mortality. The medical screening of volunteers in our study picked up these potentially fatal conditions. Since these conditions are all treatable and the risks of participating in studies with experimental drugs are small (0.04% of volunteers experience serious adverse events [1] and there are only two reported deaths [2, 3]) there may be a health benefit to certain individuals who volunteer to participate in such studies because of medical screening.

The prevalence of medical causes for not accepting subjects onto the volunteer panel in our study (10.9%) was lower than that found by Watson & Wyld [4] in their study of volunteers from the general population (14%). We also found a lower proportion of subjects unsuitable for non medical reasons than in their study (1.1% compared with 14.5%). The reason for these differences might be because our volunteers all work within the pharmaceutical industry and so would be expected to have a greater knowledge of drug development and understanding of the requirements of phase 1 studies which are intended only for healthy volunteers.

Abnormal liver function tests were found in 24 volunteers of which 22 were considered to be due to excess alcohol intake based on a history of alcohol use. None had clinical signs of liver disease or alcoholism. In the remaining two subjects the cause of liver function abnormalities was considered to be due to oral contraceptive use in one case and previous infective hepatitis in the other case. A further eight volunteers with normal liver function tests were found to have an alcohol intake (obtained at medical history) above recognized safe limits [8]. Consequently 30 volunteers (2.3%) were excluded from panel membership for probable alcohol related reasons.

Screening for thyroid disease revealed hypothyroidism in 0.5% (1 in 200 subjects) while thyrotoxicosis was not found in any volunteers. In a survey of the prevalence of thyroid disease in an English town (Whickham) hyperthyroidism was reported to be more common than hypothyroidism [14]. However, the majority of these cases had been previously diagnosed and these patients would not have been expected to volunteer for membership of our volunteer panel. In the Whickham study previously undiagnosed clinical hypothyroidism (clinical features and abnormal thyroid function tests) was found in 5 out of 2779 subjects (0.2% prevalence). However, a raised thyroid stimulating hormone alone was found in an additional 7.5% of females and an additional 2.8% of males. In our study such individuals would have been diagnosed as having subclinical hypothyroidism [15] and excluded from the panel. Previously undiagnosed hyperthyroidism occurred in 0.2% of the Whickham population. Therefore, our finding that at medical screening the prevalence of previously undiagnosed hypothyroidism is more common than previously undiagnosed hyperthyroidism is similar to that found in the Whickham study overall. The incidence of hypothyroidism in our volunteer panel was 0.05% per year. In a follow up study of the Whickham population the incidence of hypothyroidism was reported at 0.35%/year for females and 0.06%/year for males [16]. The prevalence and incidence of previously undiagnosed thyroid disorders was higher in the Whickham survey which may be because they studied a more elderly population than our volunteer panel which was restricted to those aged between 18 and 62 years. Our results indicate that thyroid function testing should be an important component in the health screening of volunteers before participation in clinical research and that repeat testing is of value to protect volunteers’ safety despite this procedure not being recommended for the general population [15].

The incidence of medical conditions occurring in panel members was 0.8% per year. The most common conditions found were anaemia, cardiovascular abnormalities, vasovagal syncope, hypothyroidism, liver function test abnormalities and haematuria. Anaemia occurred in volunteers who had all donated blood for research purposes which occurred despite this unit following the Blood Transfusion Service guidelines [17] for blood donation quantities. The majority of cases occurred in female blood donors who are more at risk of iron deficiency anaemia than male donors [18]. Blood ferritin levels are now measured annually in volunteers on our volunteer panel in order to predict those who are at risk of anaemia through repeated blood donation.

Hypertension was the commonest cardiovascular abnormality found. The pre-excitation syndrome was noted in two volunteers (incidence 0.05% per year) with previously normal ECG’s and one volunteer at initial screening (prevalence 0.08%). There were no associated dysrhythmias on their 24 h ambulatory ECG recordings. These findings are consistent with other reports of the incidence of the pre-excitation syndrome which vary between 0.1 and 3.1 per thousand per annum [19]. The term pre-excitation refers to the premature activation of the ventricle by an accessory pathway [20] which may lead to paroxysmal tachycardias as the AV node and His-Purkinje systems are by-passed [19]. The most common syndrome is the Wolf Parkinson White syndrome which is characterized by a short PR interval and a slurred initiation of the QRS complex—the delta wave [20]. It is also a recognized feature of the pre-excitation syndrome that it may be intermittent [21] and Munger et al. reported that 22% of patients with Wolf Parkinson White syndrome had no features of pre-excitation on their initial ECG [22]. This suggests that cardiac screening of volunteers should certainly include serial 12-lead ECGs to detect this syndrome which is associated with paroxysmal tachycardias and sudden cardiac death [23] and can be exacerbated by drugs [24]. Guidelines for the use of 24 h ambulatory ECG monitoring in health screening and clinical studies have previously been reported [6].

Vasovagal reactions are caused by an autonomic imbalance due to vagal overstimulation [22]. This results in symptoms such as light headedness, weakness, nausea and vomiting. Hypotension, bradycardia and loss of consciousness may be observed [25, 26]. Over the 4-year period of the study 0.13% of volunteers per year experienced vasovagal episodes during phlebotomy which were severe enough for them to be excluded from further participation in the panel. Less severe vasovagal episodes were experienced by other volunteers which did not require exclusion from the panel but the incidence of these was not quantified in our study. The prevalence of all vasovagal reactions in blood donors has previously been reported as 2.1% including 0.1% regarded as severe [27]. Cardiac arrest following venepuncture in a healthy subject has also been described [28]. It is important to identify individuals who suffer with vasovagal episodes during medical procedures as early studies of experimental drugs with unknown cardiac effects in humans could be potentially hazardous for such volunteers, e.g. if the drug supressed sinus node or conducting pathway activity or acted as a vasodilator [29].

In summary, this study demonstrated the importance of performing medical screening before volunteers are allowed to participate in clinical research. In our unit previously undiagnosed medical conditions were the most common reason for exclusion from the healthy volunteer panel. Additionally, repeating the screening of subjects enrolled on to the panel at regular intervals identifies previously ‘healthy’ volunteers who have developed new medical conditions.