Chapter 1 – Introduction and objectives

Employing over 58,000 people worldwide, AstraZeneca is a leading pharmaceutical company with sales in over 100 countries, manufacturing in twenty, and major research centres in five.  AstraZeneca was formed on 6 April 1999 through the merger of Astra AB of Sweden and Zeneca Group PLC of the UK – two companies with similar science-based cultures and a shared vision of the pharmaceutical industry.  Sales in 2002 totalled $17.8 billion, with an operating profit of $4.4 billion [4] . Societal expectations relating to corporate responsibility are higher than ever before.  Maintaining society's trust and confidence is an increasing challenge for companies everywhere.  Shareholders are of course still looking for a good return on their investment, but they increasingly want assurance that they are investing in a business that delivers shareholder value in a responsible way.  Customers expect high ethical standards as well as high quality products.  Local communities only welcome those businesses that can demonstrate good corporate citizenship.  Effective corporate responsibility performance depends on achieving an acceptable balance between the economic, environmental and social priorities of sustainable development.  Quoting AstraZeneca Chief Executive Sir Tom McKillop [4] :  

"Corporate Responsibility is not an add on extra - it is an integral part of all that we do and I am determined that we will continue to be a company that is welcomed as a valued member of the global community”.  

Safety, health and environmental (SHE) performance is key to building and maintaining effective corporate responsibility.  As regulatory frameworks become ever tighter and the cost of business interruptions soars, there is an increasing need for continuous improvement in SHE compliance based on a creative and self-fulfilling compliance culture.  The pharmaceutical and other related process industries are characterised by inherently hazardous processes and activities.  Besides the obvious areas of safety, health and the environment, organisational decisions and assessments based upon relative risk are made frequently and include:    

The challenge for most organisations is how to identify where the greatest SHE risks lie.  A number of risk profile identification strategies have been used by industry, including: 

Despite the desire to improve SHE performance, accident and incident rates for many commercial and industrial organisations have plateaued [48, 99].  The reason behind the plateau is perhaps that organisations have historically looked toward hardware and procedural control to better improve SHE performance and compliance.  Saari [143] suggested that after a certain point, technology alone cannot achieve further improvements in safety.  Instead, organisational and cultural factors become more important.  Awareness regarding the importance of organisational culture and its relationship to positive SHE performance is growing [59, 122]. 

More and more organisations are realising that hardware and procedural improvements do not, by themselves, give rise to tolerable levels of compliance [96] .   Assessments have been made that attribute over 50% of industrial-based accidents to factors such as poor management, poor training and other psychological factors [97] .  Other studies within the chemical industry have shown that more accidents are attributable to human issues and managerial deficiencies than to weaknesses in technical components.  For instance, Powell and Canter [132] , Reason [138] and Layfield [102] highlighted that recognising the contribution of humans to failure of systems has long been an integral part of good engineering design. 

Engineers’ attempts to quantify the frequency of human failures are generally based upon historical data and are limited to low order human tasks such as, for example, ‘wrong button pressed’, ‘material added in wrong order’ and ‘processing step missed out’.  These foreseeable deviations from routine operating conditions are often identified using processes such as hazard and operability studies (HAZOP) [98] .  Higher order management failures such as ‘inappropriate purchasing strategies’ and ‘deliberate actions’ are often excluded from these studies. 

Authors such as Heinrich et al [73] and Porter and Corlett [131] consider that more than 90% of accidents are caused by human factors rather than equipment failures.  Powell and Canter [132] found that within the chemical industry, more accidents are attributable to deficiencies in human and management components than to unforeseeable weaknesses in technical components.

The findings of major accident investigations are showing that human cultural factors play an important role in accident causation, prevention and mitigation.  Companies are realising that ‘rule based’ organisational cultures do not adequately ensure satisfactory performance.  Two major reasons are identified:  firstly, no one carries a rulebook around with them.  Secondly, it is not possible to write rules that encompass all activities and circumstances.  Recognising that personnel in all jobs have, to an extent, degrees of interpretive flexibility, the goal of industry should be to ensure that personnel have internalised, via training, their perceptions and the messages that they have received about behaviour that is or is not acceptable.  The challenge for highly regulated industries is to find the balance between rule setting and allowing sufficient flexibility to enable personnel to work effectively and efficiently.   

  There are several possible explanations why industry has generally failed to address human factor issues.  The most likely explanation is perhaps that hardware plant and equipment is more tangible (therefore easier to address) than ‘human’ and ‘cultural’ factors.  Another reason for the failure may be due to a requirement for organisations to reach a level of engineering and procedural maturity before it becomes realistic for them to tackle human factor issues.

The importance of addressing cultural aspects has been highlighted by recent well publicised major loss events such as Chernobyl [67] , Piper Alpha and the Kings Cross fire.  Following the 1988 Kings Cross Fire, Judge Fennell [54] wrote:

  “... a cultural change in management is required throughout the organisation”.

Similarly during the Piper Alpha Inquiry Lord Cullen [32] wrote:  

“ … it is essential to create a corporate atmosphere or culture in which safety is understood to be and is accepted as, the number one priority”.

More recently, Lord Cullen’s [33] inquiry into the 1999 Ladbroke Grove rail accident paid much attention to the role of safety management and culture within the rail industry.   The inquiry pointed to evidence that suggests that a large proportion of accidents, incidents and near miss occurrences follow from unsafe acts that are a result of underlying deficiencies in safety management.  Thus, the inquiry emphasised the link between ‘good safety’ and ‘good business’.  Cullen recognised that a successful safety culture depends on leadership.  The inquiry heard that the fragmentation of the rail industry made it difficult to achieve clear safety leadership within the UK rail industry.  Cullen stressed the need for the safety commitment of senior management that should be clearly visible to front line workers and, the need for effective communication of safety goals and objectives along with regular meetings devoted to safety issues.  Dyer [51] , Greenstreet Berman Ltd. [66] and the HSE [86] have summarized those good practices that support a positive SHE culture. 

The recognition that there is a relationship between organisational culture and safety performance has spawned an increased interest in identification of methods that allow measurement of organisational culture [68, 88, 154, 177].  Organisational culture can be measured and evaluated through organisational climate surveys.  Organisational climate surveys can be viewed as surface indicators of an organisation’s underlying culture.  Culture is a multi-dimensional construct.  No one dimension, factor or attribute can therefore adequately describe it.  Safety culture is only one of the many dimensions of an organisation’s multi-dimensional culture.

A significant amount of research associated with organisational climate surveys exists.  These surveys tend to be almost entirely focused toward issues that can be directly associated with safety.  The surveys are generally composed of a set of questions that typically require a ‘tick box’ preference response from ‘strongly agree’ through ‘neutral’ to ‘strongly disagree’.  Researchers typically group survey questions together under hypothesised factors.  The response data are often subjected to structural equation modelling techniques to identify the complex relationships between each of the hypothesised factors.

Surveying a number of cultural surveys available in the literature indicated that the number of principal component factors of organisational culture identified ranged from two [36] to nineteen [104] .  In addition to variation of the number of factor components, the literature survey indicated that the components themselves varied from survey to survey.  Flin et al [60] carried out a thematic survey of eighteen cultural surveys and concluded that all of the themes could be distilled into the following five factors:   

Unsurprisingly, as is the case with many social science issues, no agreement currently exists as to how many factors or indeed which factors are required to adequately describe organisational culture.  Several studies have examined the results of attitudinal surveys and have proposed architectures or underlying structures of the shared perceptions and attitudes toward safety [36, 88, 163].  The models often suggest strengths of relationships and interdependencies of the measured cultural factors. 

Although much work has been done regarding the identification and modelling of safety culture factors, significantly less research has been performed in correlating theoretical cultural constructs with broader outcome metrics such as lagging safety performance indicators.  A study by HSE [88] found statistically significant relationships between cultural factors and lagging safety indicators such as self-reported injuries and dangerous occurrences.  Coyle et al [31] found a relationship between safety climate factor scores and lost time accident rates, although this relationship was not statistically quantified.  Hofman and Stetzer [76] found that safety metrics were related to accident rates.  These studies are similar to the majority reviewed in that they focus upon those organisational aspects that are directly related to safety.  Little research was found that has examined how other non-safety related organisational factors may affect SHE performance. 

In response to these findings, AstraZeneca has started to take a more holistic approach to SHE management in which all the contributing factors to potentially unsafe incidents are considered.  This holistic approach requires examination and development of a range of issues including the traditional safety technology and engineering controls together with management systems and human factor issues.  

AstraZeneca recognises that employees’ attitudes and satisfactions can strongly influence its ability to meet business objectives.  In an attempt to measure employee attitudes and behaviours the company has administered three ‘Focus’ surveys.  The first survey was conducted in 2000, the second in 2002, the third in 2004.  All AstraZeneca sites took part in each of the surveys.  About 42,000 personnel responded to the 2002 survey, making it 5 times larger than the next largest comparable survey found in the literature review [154] .   The Focus surveys were used to measure employees’ attitudes and satisfactions in ten factor areas.  Comparison of the Focus 2002 survey questions with established organisational culture research indicated that the Focus data set would be a useful and effective measurement tool that would allow AstraZeneca to measure its organisational culture.

It is imperative that any safety performance indicators used to correlate with organisational climate metrics are comparable and robust.  A sufficient number of sites must have comparable, if not identical, metrics for the correlational exercise to be valid.  The lack of a consistent set of organisational climate and performance metrics across industry is probably the principal reason for the lack of research in this area.  Despite their potential usefulness, organisational climate metrics have not yet been exploited by industry as a proactive measure of SHE performance or as an aid to relative risk ranking.

As part of AstraZeneca corporate procedures, each site within the company is required to report its SHE performance on a quarterly basis.  Guidance is given within the corporate reporting procedure to ensure consistent reporting. 

The availability of the Focus 2002 survey and SHE performance indicators covering several countries, presents a unique opportunity to explore the links between organisational culture and SHE performance.  The research objectives listed in Table 1 were formulated to address this opportunity.

Table 1.1 (continued) Research questions

  ^* Answers to these research questions will contribute to existing knowledge.  

Answering the above questions will enable industry to determine the usefulness of organisational culture metrics to predict SHE performance and allow identification of those organisational culture factors that most influence it.