Editorial |
COPD and the workplace. Is it really possible to detect early cases?
Chronic obstructive pulmonary disease (COPD) remains a significant personal and public health issue. The disease is normally characterized by a slow but progressive decline in lung function leading in some to eventual respiratory failure and death. The recent announcement that a National Service Framework should be developed to deal with various medical aspects of this condition is recognition of the damaging effects this condition has on society.The epidemiology of COPD suggests strongly that this is largely a preventable condition. Cigarette smoking is the most important and removable cause of COPD. While other risk factors are consistently identified, research mainly relates to diagnostic and treatment strategies with little regard to identifying workers at risk of this condition by virtue of their harmful workplace exposures.
However, various recent studies have assessed the effect of workplace exposures on COPD, and reveal that a proportion of the total ‘COPD burden’ is caused by airborne occupational exposures, in addition to cigarette smoking. This is not a new finding. The potential link between workplace exposures and respiratory ill-health dates at least to the 15th century, although it was Ramazzini in the 17th century followed by Greenhow [1] and others who popularized this connection. The magnitude of this effect is now known to be in the order of 15% of all COPD, also termed the population attributable risk fraction [2]. That is, approximately one-eighth of the total societal burden of this devastating illness would not have occurred in the absence of inhaled harmful workplace exposures.
Unfortunately, recent studies generally lack accurate population exposure assessment, information that would enable us to answer the important question ‘which agents and working practices pose the greatest risks?’. In addition, how these potentially harmful workplace exposures interact with smoking (and other potential risk factors such as genetic predisposition) is not fully understood, although it is likely that all potentially harmful agents interact in a complex fashion to produce lung damage.
In order to reduce the substantial societal burden related to COPD over the next few decades, it is essential to advocate smoking cessation and to identify workers who are at risk of developing this disease, irrespective of the effect of cigarette smoking. So important is this issue that the Health & Safety Executive has identified work-related COPD as a major component of its current disease reduction programme. This inclusion will hopefully identify key groups with interest in reducing the substantial burden of occupational COPD, and allow educational and research programmes to be developed to assist the desired reduction.
Once a risk arising from a specific exposure is clearly established, occupational health professionals are ideally placed to further implement this important task, and should grasp this responsibility firmly. While on the surface this may seem fairly straightforward, there are a number of significant hurdles to be cleared.
These include answering important questions such as can COPD, including work-related COPD, be detected early (for example by exaggeration of the normal decline in lung function with age) such that eventual diagnosis can be predicted and preventive action taken? Similarly, can rate of decline be measured sufficiently accurately in workplace settings to be of use in guiding individuals about continuing employment? Again, can evidence of risk be detected in workforces before the clinical condition can be detected in individuals and consequently can workplace surveillance provide evidence of the effectiveness of control measures?
In order to identify workers ‘developing’ COPD who do not already meet the diagnostic criteria, it is not currently sufficient to rely on reported symptoms in the workplace. Symptoms of COPD (breathlessness, cough, phlegm production, chest tightness and wheeze) are common in those without COPD, and it is not uncommon to identify workers without symptoms who satisfy the diagnostic spirometric criteria. Symptoms changing over time, therefore, are not sufficiently sensitive or specific to pick up early or ‘at-risk’ cases. Symptoms should not be ignored in this context, however, and may identify an at-risk population for developing progressive airflow obstruction.
So, key to the diagnosis of COPD are measures of lung function, interpreted with reported symptoms. COPD is defined physiologically by the presence of airways obstruction (FEV1/FVC ratio <0.7), with an accompanying absolute reduction in the FEV1 to <80% of its predicted value [3], although there are subtle differences between varying definitions.
Single measures of lung function therefore enable the occupational physician to define the presence and extent of COPD physiologically and allow comparison of the measured FEV1 and FVC with predicted values. Such ‘cross-sectional’ evaluation of lung function, using reference lower limits of normality, allows absolute values of lung function to be observed instantaneously without waiting for significant drops in lung function over time.
Such abnormalities of lung function are commonplace in the general population [4], and consequently also in working populations. For example, recent data from our unit assessing flour-exposed workers show that of 483 workers assessed, 26 (or 5.4%) met the physiological criteria for COPD. From this information, however, it is impossible to assess the rate of lung function decline leading up to the measurement point.
Consequently, measures of FEV1 and FVC are best taken over time, in order to compare values and calculate the rate of decline in lung function values. Reliable values for rate of decline in ‘normal’ non-smokers and smokers exist, and serve as a good comparison for exposed workers.
While this exercise may seem attractive to occupational health professionals, there are reasons why this approach is probably not commonplace. The spirometer measuring FEV1 and FVC may itself be a significant source of error. For example, if a spirometer is accurate to ±3% [the current American Thoracic Society (ATS)/European Respiratory Society requirement] [5] and a worker has a truly stable FEV1 on two occasions, 1 year apart, of 3.5 l, the spirometer could return a reading of +3% at baseline (3.61 l) and conceivably return a reading of –3% the following year (3.4 l). This amounts to a hypothetical measured annual decline of 210 ml, approximately seven times the expected loss of 30 ml per year for a non-smoker [6]. This calculation also assumes that all other measurement-related factors are as precise as possible.
It is therefore clear that much larger falls in FEV1 are ‘required’ for occupational health professionals to be sure that they are significant or greater than would be expected. Therein lies the current problem, magnified when dealing with groups of workers, where the variation of measures within individuals over time and between individuals is not known.
Various recent publications have addressed this issue. Hnizdo et al. [7] recently suggested in their excellent study using data from 11 US workplaces that precision of longitudinal measures of FEV1 was a critically important factor in picking up early cases of rapid annual FEV1 decline. A greater degree of precision in measures taken over time would allow earlier identification of workers with rapidly falling FEV1 values.
The example cited by this paper calculated that when assessing decline in a 34-year-old man, high precision of measurements over each year (as defined by a low workplace-derived mean within person standard deviation of pairwise FEV1 measures) would identify significant FEV1 loss 4 years earlier than if measures were taken with low precision.
Using information from this paper, Hnizdo estimates with a given degree of precision what levels of decline would be regarded as significant, and the outcome is surprising. The authors suggest that >1 year, declines in FEV1 of <330 ml would not be easy to attribute to ‘significant’ loss, if measures were accurately recorded (with a within-person variation of 130 ml). The loss per year that can be significantly identified reduces as one might expect with longer duration of follow-up and greater measurement precision. With a follow-up length as generally recommended for valid interpretation of serial spirometric data (i.e. 4–6 years), excess annual declines of the order of 120–150 ml may be detected, assuming two repeat measures are taken with the same level of precision.
These figures agree well with the trigger level suggested by the ATS [8] and NIOSH [9]. Both regard a 15% annual excess decline in FEV1 as an appropriate trigger level for defining a clinically significant decline in lung function. Indeed, for a 45-year-old male of average height with a predicted FEV1 of 3.64 l, a 15% decline is equivalent to an absolute volume of 
550 ml. Interestingly, recent data from an epidemiological investigation of a large cohort of coal miners suggest that this approximate loss, albeit from cross-sectional data, is associated with a doubling in the risk of MRC grade-3 breathlessness [10].
So where does all this leave the practising occupational health professional responsible for assessing workers' lung function? A consensus of opinion appears to be emerging in this complex area.
First, if a decision is made to measure FEV1 and FVC in the workplace (and this would be regarded as good practice for all workers at significant risk of either asthma or COPD), this should be measured to appropriate standards. Lack of precision (a function of many factors including inadequate operator training, inadequate machine calibration, multiple operators over time) seriously undermines the use of this test for identifying workers losing FEV1 more quickly than they should, and can delay this process by years, in turn delaying appropriate medical assessment and intervention. All those performing spirometry in the workplace should have received specific training.
Second, minor measured changes in excess of those expected over 1 year should probably be ‘ignored’, but documented. In cases of doubt, adding further accurate measures of lung function at earlier time points may help assess a more accurate decline rate for FEV1. While less mathematical, it would seem reasonable to remeasure FEV1 after 6 months in those with a previous drop of 250 ml or more over the previous year, although access to appropriate testing may pose local difficulties.
Third, a fall in FEV1 of 15% from baseline over 1 year (or of 500 ml over 1 year), or changes of 100 ml per year for 5 years, should certainly ring alarm bells, although some might argue that waiting 5 years to confirm this is difficult to justify, not least to the worker themselves. These workers need careful counselling, and referral to a specialist with an interest in occupational lung disease.
The decision to intervene in the workplace, and to prevent an early case of airflow obstruction causing disability, is complex and must always take into account information on reported symptoms, smoking habit and workplace exposures. This difficult challenge largely lies before all of us with medical responsibility for working populations, and is clearly a key role for all occupational health professionals.
Reader and Honorary Consultant Respiratory Physician, Centre for Workplace Health, University of Sheffield
Senior Scientist, Centre for Workplace Health, Health and Safety Laboratory
References
- Greenhow EH. (1861) Report of the Medical Officer of the Privy Council. Appendix VI(HM Stationary Office, London).
- American thoracic society statement: occupational contribution to the burden of airway disease. Am J Respir Crit Care Med (2003) 167:787–797.
[Free Full Text] - National Institute for Clinical Excellence Management of Chronic Obstructive Pulmonary Disease in Adults in Primary and Secondary Care.NICE Clinical Guideline 12. 2004.
- Halpin DMG and Miravitlles M. (2006) Chronic obstructive pulmonary disease. The disease and its burden to society. Proc Am Thorac Soc 3:619–623.
[Abstract/Free Full Text] - Miller MR, Hankinson J, Brusasco J. (2005) Standardisation of spirometry. ATS/ERS taskforce: standardisation of lung function testing. Eur Respir J 26:319–338.
[Abstract/Free Full Text] - Hankinson JL and Wagner GR. (1993) Medical screening using periodic spirometry for detection of chronic lung disease, state of the art reviews. Occup Med (Lond) 8:353–361.
- Hnizdo E, Yu L, Freyder L, Attfield M, Lefante J, Glindmeyer HW. (2005) The precision of longitudinal lung function measurements: monitoring and interpretation. Occup Environ Med 62:695–701.
[Abstract/Free Full Text] - American Thoracic Society. (1991) Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis 144:1202–1218.[Web of Science][Medline]
- Criteria for a Recommended Standard: Occupational Exposure to Respirable Coal Mine Dust September 1995 NIOSH Criteria Documents. DHHS (NIOSH) Publication No. 95–106.
- Cowie HA, Miller BG, Rawbone RG, Soutar CA. (2006) Dust related risks of clinically relevant lung functional deficits. Occup Environ Med 63:320–325.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||