Communication strategies for product recalls in the healthcare business

Table of contents

List of figures

List of tables

Glossary

References

Abstract

1 Introduction
1.1 Research purpose and research questions
1.2 Limitations
1.3 Structure of the thesis

2 Literature review
2.1 Attribution
2.2 Perception
2.3 Communication channels
2.4 Crisis and reputation
2.5 Legal aspects
2.6 Learning from a crisis
2.7 High reliability and normal accident theorists
2.8 The business impact of bad news
2.9 Stakeholder analysis
2.10 Corporate communication models

3 Research methodology
3.1 Research strategy and design
3.2 Data sources
3.3 Analysis techniques
3.4 Ethical implications of the research

4 Findings
4.1 Product recall analysis
4.2 Communication response to the recalls from different stakeholders
4.3 Stakeholder analysis

5 Discussion and interpretation
5.1 General discussion of the findings
5.2 Communication strategies of the companies under study
5.3 General discussion of trends
5.4 Who drives the discussion in the healthcare community
5.5 Methodological aspects and validity

6 Conclusions and recommendations
6.1 Research questions
6.2 General conclusions and recommendations from the corporate perspective

7 Appendix
7.1 HRS FAQ on Policy Recommendations: Implantable Cardioverter Defibrillator and Pacemaker Performance
7.2 Device recall list

List of figures

Figure 1: Stakeholder mapping - the power interest matrix

Figure 2: Brief research overview

Figure 3: Class I and II recalls from all manufacturers over time

Figure 4: Guidant (now BSX) stock prices and recalls

Figure 5: BSX stock prices and recalls

Figure 6: MDT stock prices and recalls

Figure 7: ELA stock prices and recalls

Figure 8: STJ stock prices and recalls

Figure 9: Detailed view on Guidant's (now BSX) recall on 30.6.05 and 23.11.05

Figure 10: Detailed view on BSX's recall on 28.9.06

Figure 11: Detailed view on ELA's recall on 2.12.05

Figure 12: Power-interest matrix in recall processes

List of tables

Table 1: Number of recalls per class per company

Table 2: Number of different reasons for recalls per company

Table 3: BSX recalls over time with class, number reasons and devices

Table 4: MDT recalls over time with class, number reasons and devices

Table 5: ELA recalls over time with class, number reasons and devices

Table 6: STJ recalls over time with class, number reasons and devices

Table 7: Company response to recalls (unusual behaviour in bold)

Glossary

illustration not visible in this excerpt

References

illustration not visible in this excerpt

Web references

illustration not visible in this excerpt

Abstract

Ashcroft International Business School

Master of Business Administration

Communication strategies for product recalls in the healthcare business

By Thomas Osche

August 2007

After a series of cardiac implantable device recalls hit the industry in the year 2005 the question of a favourable corporate communication strategy about inevitable product recalls was raised. Knowing that it is impossible to produce 100% failure free products I was personally interested in seeing how such recalls occurred and how successful they were handled by the different companies from a communication perspective. This thesis analyses such device recalls since the year 2000 and attempts to identify a favourable communication strategy for the companies involved in such recalls in the future.

Qualitative analysis was performed on various secondary data sources to find positive and negative aspects of different communication strategies for the affected companies. Four companies could be analysed, where three provided significant results, which were used for final recommendations. Those recommendations are provided to support active corporate communication. The basis of the recommended communication strategy is as follows:

Provide disclosure about probabilities of malfunctions while supportive to the most powerful stakeholders, which are the physicians and the regulatory authorities. When one company faces a recall it is favourable to take accountability and give advice to physicians about how to respond to their patients, because they are facing the problem of having all the information at hand without an adequate guideline which summarises a clinical decision on how to act and how to weight risks of keeping the existing device versus the risks of device change-out procedures.

Keywords: communication strategy, crisis management, corporate communication, healthcare business

1 Introduction

This master thesis focuses on the implantable medical devices area of the healthcare business. There are five noticeable companies working in this area worldwide. Three of those are American companies and two are European. Implantable medical devices, or more precisely implantable pacemakers or cardioverter defibrillators, are used if medical treatments are no longer sufficient for a patient’s pathological health condition. Pacemakers help patients if they have problems with a slow heart beat or chronothope incompetence, and a cardioverter defibrillators is indicated for patients with a risk of sudden cardiac death, which can be cause by extremely fast and irregular heart beat contractions.

Thus those devices are either permanently supporting the life of a patient or they save the life at a certain critical incident. This research is concerned with the communication strategies of the companies with respect to device problems, or recalls. The term recall does not necessarily mean the device is recalled back to the company. It means that the product which is already officially available on the market has a problem which requires special advisory from the company to the physician or the patients. Sometimes this means that special procedures should be followed in order to provide a safe function of the device but sometimes it really means that a device has to be changed.

Product recalls are an expensive business process that needs to be managed carefully. The negative impact on patient’s safety as well as on costs needs to be minimized. Companies communicate about product recall with the public as well as with the regulatory authorities; hence it requires intensive planning and coordination.

On 22 July 2005 the Guidant Corporation, now owned by Boston Scientific announced a recall of several pacemakers. This recall consisted initially of a letter to physician that described the problem and provided recommendations to minimize the risk of a pacemaker failure. The following summary of the recall facts is based on [FDA2].

The FDA, as the regulatory authority for the American health care system, has classified Guidant's action (letter to physician) as a Class I recall. Recall classifications can fall into one of three categories, with Class I being the most serious. These numerical classifications are based on the probability that the device failure could lead to adverse health effects. In a Class I recall, there is a reasonable probability that the malfunctioning device will cause serious adverse health consequences or death.

"Pacemakers are complex medical devices that can extend and improve the lives of many people who have heart rate abnormalities. However, they are not perfect and can malfunction," said Daniel Schultz, M.D., Director, FDA’s Center for Devices and Radiological Health. "We are notifying patients and physicians about this important safety matter so they can take prompt action to reduce the risk of serious health consequences."

This recall was initiated after 69 reports where filed with Guidant, that indicated a leakage problem in several device-types of the company. Approximately 18,000 of the affected devices remain in service in the United States and an additional 10,000 are in service in other countries. Guidant estimated that the failure rate from the leakage defect lies between 0.17% and 0.51% (i.e., between 1.7 per one thousand and 5.1 per one thousand) over the remaining lifetime of the devices.

The FDA did not made a recommendation about whether a patient who had one of the Guidant pacemakers affected by this recall should have it replaced. This is a decision that should be made by the patient in consultation with his or her physician, based on the patient's history and medical condition. Removal and replacement of the device may pose some risk, so it is important that patients and physicians carefully discuss this matter before making a decision.

Some of the recommendations to the patient are as follows:

- If you believe you are pacemaker dependent, contact your physician soon to discuss your treatment options.
- Continue your normal doctor appointments.
- If you experience symptoms of shortness of breath, dizziness, lightheadedness, loss of consciousness, or a prolonged fast heart rate, you should consult with your physician or go to the emergency room immediately.

This recall was accompanied not only with a FDA class I recall announcement, the letter to physician, recommendation to patients, a press release, extended news coverage due to the magnitude and severity of the problem, information websites and hotlines for physicians, patients, and relatives, but also with a series of previous and following recall events by the Guidant Corporation and the other companies. This recall eventually leads to stronger regulation and control of the companies by the FDA, and to an increased awareness of patients induced by profit oriented lawyers. Both of these consequences seem to have had a major effect on the whole business with implantable devices, towards increased safety and liability awareness.

A product recall in the implantable device industry can have different reasons such as defect components from supplier or faulty product behaviour with the root cause in the company’s internal processes and quality inspections.

The objective of this research is to identify favourable communication strategies in the field of contingency and crisis management, specialized to the healthcare industry.

Brands can have a complex communication system of communication channels. The proposed research will analyse which communication channels are active, inactive or have a changing utilization before, during and after a product recall. Additionally the communicated content will be analysed. Finally the recall communication strategies of different companies or different brands with recall incidents will be deduced and analysed with respect to their success in terms of business or brand impacts. This analysis will therefore lead to recommendations for the communication strategy of a company working in this area of the healthcare industry.

1.1 Research purpose and research questions

This research aims to find recommendations for the corporate communication strategies of companies that deal with implantable medical devices. The research questions (RQ) are as follows:

RQ1: What are the communications strategies of companies that deal with product recalls of implantable devices?

RQ2: What makes those strategies successful?

1.2 Limitations

This research focuses on a small area of implantable devices, namely pacemakers and defibrillators. The nature of those devices is to save the patient’s life, this makes them special and this research can only provide insight in this specific area of healthcare. Furthermore this thesis relies on secondary research of qualitative sources. It will of course analyse multiple cases of recalls, but still depends on the possibilities to generalize the issues within those cases. Further details on limitations are provided with the research methodology in chapter 3.

1.3 Structure of the thesis

After a literature review in chapter 2 on communication, perception, reasons for product failures, and other relevant areas, the research methodology will be elaborated in chapter 3. In chapter 4 the findings of this research are presented objectively and focused on facts only, followed by a discussion and interpretation in chapter 5. This separation of findings and interpretation will facilitate the potential generation of a crisis communication model in the business area under investigation. In chapter 6 conclusions and recommendations will be presented to contribute to the ongoing discussion of the handling of device failures in the healthcare business.

2 Literature review

Communication strategies are effective if they have the following characteristics: They need to be consistent, adaptive to the environment, maintain the competitive advantage and be feasible [Varey2001].

The communication type that needs to be analysed in detail in this research is external crisis communication. The differences in the approaches to handle and organize those communications can be striking. In many cases the crisis communication even has the opposite effect to the intended meaning. Szukala2001 sees the difficulties especially in the art of impression management, because there are so many uncontrollable factors that affect the observers’ perceptions.

Because impression and perception are very important issues in this research this chapter starts with review on attribution and perception literature.

2.1 Attribution

The theory of attribution [Weiner1986] helps to understand consumer reactions to recalls or disasters. They are related to three dimensions:

- The locus – the extent to which the cause of the disaster was internal or external to the organization concerned
- The controllability – how far the cause was within the control of the organization
- The stability – whether the disaster’s trigger is predictable or subject to constant change.

Attribution theory assumes that people try to determine why people do what they do, which means that they try to find causes to an event or behaviour. A three-stage process underlies an attribution:

- behaviour must be observed/perceived
- behaviour must be determined to be intentional
- behaviour attributed to internal or external causes

When one succeeds, one attributes successes internally (my own skill). When a rival succeeds, one tends to credit external (e.g. luck). When one fails or makes mistakes, we will more likely use external attribution, attributing causes to situational factors rather than blaming ourselves. When others fail or make mistakes, internal attribution is often used, saying it is due to their internal personality factors.

1. Attribution is a three stage process: (1) behaviour is observed, (2) behaviour is determined to be deliberate, and (3) behaviour is attributed to internal or external causes.
2. Achievement can be attributed to (1) effort, (2) ability, (3) level of task difficulty, or (4) luck.
3. Causal dimensions of behaviour are (1) locus of control, (2) stability, and (3) controllability.

Weiner’s attribution theory is not only interesting when analysing reaction to crisis or disasters it is also interesting when one analyses multiple text sources. If one researcher (that would be me in this research) reads multiple text cases he has to admit that he is not completely free of preconceptions, thus has a probability of false interpretation or false attribution. These preconceptions need to be minimized in order to analyse the data as objectively as possible. Weiner’s theory helps not only to understand the data under research but it helps also to understand some of my own mechanisms of thinking and interpreting data and therefore helps to provide an objective research.

2.2 Perception

According to Regester2005 p.163ff a crisis itself contains seeds for success as well as the roots of failure. Applied to the area of recalls of devices his statements lead to the insight that it is very important to be seen to take action in case of a crisis and to be heard to say the right thing. The perception you can create for your customers is of utmost importance. He states that companies often focus too much on technical aspects and ignore the issues of perception. This will be an important aspect of the data analysis, as it may show differences or similarities between the different recalls of the different companies. The perception can be anticipated from the media usage of the companies, where BSU2007 p.63ff (chapter: Analysing the communication situation) provides a communication model, which should provide insight into the information flows.

The model combines the following components:

- stimulus
- (pure) perception
- the act of memorizing
- preconscious perception
- cognition
- personal cognitive structures
- intended actions
- observable behaviour

But in this research I can only hope to find information about stimulus and the observable behaviour. The stimulus is a complex message which comes from the message sent and its surroundings. Which part of the message is actually perceived not only depends on the message itself, but also on the content of the messages perceived before. Pure perception means in this model only the physiological perception, which is of no interest for this study and should not be confused with the broader, more casual meaning of the word. The other components are in their detail of minor interest to this study, but it is worth to know that they are there, because they form the broader understanding of (not only pure) perception. The final component is the observable behaviour which does not require any explanation.

In the context of this model it is argued in BSU2007 p.63ff, that marketers need to know exactly who uses media or media vehicles and when and how is this media used. They need to know this because they need to know what state of mind their receptors have and how they have been approached. So what has to be done with this research is not really different from what marketers or media analysts do.

2.3 Communication channels

The communication channels used in the healthcare business are not classic consumer channels but specific business-to-business channels, like journals, conference papers or presentations and direct talks between the sales representatives and the doctors and the hospital managers. Conference papers like RD2005, Ludgate2006 or Nisam2006 should be treated with some carefulness, because they could be influenced by companies, even if they appear to be objective.

In Argenti’s corporate communication model (see chapter 2.10.1) the following rather classic channels are mentioned in order to communicate: corporate advertising, media relations, internal communication, investor relations, government relations and crisis communication. The analysis of the corporate communication in the healthcare business should of course analyse those channels but also should especially account for the healthcare specific channels mentioned above.

2.4 Crisis and reputation

A very interesting article from Coombs2004 analyses the impact of multiple crises communication to the reputation of a company and also analyses the correlation between reputation and the perception of the responsibility for the crisis. But before I present his system let me very briefly analyse what reputation is and how it is generated: Reputation is built because people seek for trust in relationships. Reputation is built over time and is also built because the experience with things or people was useful or cheerful. In other words: The main reason that we read the Wall Street Journal is that we’ve found it useful in the past. (SV1999).

As we now have an idea what reputation can be alike, let’s see what Coomb2004 has analysed: He discusses a theory based system which he named SITUATIONAL CRISIS COMMUNICATION THEORY (SCCT) for matching crisis response to the crisis situation. The basis for this theory is the attribution theories (see Weiner1986). The SCCT argues that information about past crises is a significant factor that can affect perceptions of a more recent crisis. His theory will be of much help to reflect on the findings of the qualitative analysis. Two aspects of his analysis framework are of special importance for this research, because they describe especially technical, human and organizational roots for crises: The Accidental Crisis Cluster (technical roots) and the Intentional Crisis Cluster (human and organizational roots).

2.5 Legal aspects

Another aspect which is important to the field of healthcare in the event of a product recall is the legal aspect. According to Regester2005 p.186ff lawyers tend to admit nothing where an open communication may be required. So it will be researched when the companies admitted which detail about the device recalls and how the timeline of communication about a recalls does look like. According to Hübner2007 p.98ff regulations help companies to communicate because they limit the discussion between lawyers and managers to a smaller area, which effectually leads to reduced stress between those stakeholders. The healthcare business is highly regulated and regulations even increased over the time, so the analysis of legally required information is important and will be performed in the research with respect to the compliance of the companies to the FDA regulations.

2.6 Learning from a crisis

But how does a crisis arise anyway? According to ESG2000 managers have difficulties to deal with failures and also companies have difficulties to learn from failures. Some reasons for this are as follows (Only a few examples are given here): (1) The roots for the crisis are too complex. (2) Inadequate regulation and enforcement or (3) a poor problem definition leads to a crises and sometimes prevents companies from learning from own failures or failures of others.

The roots for each recall will therefore be analysed to try to find similarities or specialities in the healthcare business. For example the FDA implemented much tighter regulation in the medical device industry after a series of recalls since 2004 in the USA.

2.7 High reliability and normal accident theorists

The debate between reliability theorists and normal accident theorists introduces us into the area of conflict between ethics and technical feasibility [Lagadeck1997], which play an important role in the healthcare business, where ethical commission are common use to decide borderline cases. Reliability theorists explain that organization can have high levels of safety by having priorities on safety and redundancy and cultivating a reliability culture [Rochlin1993]. The normal accident theorists argue that it is illusory to completely avoid accidents because the organization and technical systems are of increasing complexity which leads to unexpected interactions. They also argue that safety is far from being a priority, even if stated otherwise, because the area of conflict between business success and good governance is far from being resolved. They argue that redundancies add more complexity to a system and hence have a distorting effect on reliability [Sagan1993]. This debate is especially interesting because it is an old discussion, which comes from a completely different field of business – Sagan1993 for example discusses the safety of weapons. The reference to this classic discussion is given here because a similar discussion has just started in recent years in the medical device industry and the debate about ethical standards and technical feasibility is ongoing in the healthcare business [Ludgate2006, Nisam2006]. Due to the fact that this discussion is very new, only the relation to other industry sectors was given here, but the debate in the healthcare business will be detailed and related to the classic discussion later in the research.

2.8 The business impact of bad news

Recalls are of course bad news. In EKT2005 the impact of bad and good news to the healthcare sector was analysed and it was found out, that negative news have a much stronger negative impact than good news having a positive impact. They analysed stock prices and transformed good and bad news into a domain where they could find correlations between the stock prices and that news. In their analysis all three major USA medical device companies, which are also part of this research, where included.

2.9 Stakeholder analysis

To analyse the stakeholders and their importance and consideration in the companies’ communication, Johnsen and Scholes (JS2006 p. 182) provide a very helpful tool with the power-interest matrix. This diagram classifies or maps stakeholders into the matrix and denotes their importance or general behaviour a company should show to those stakeholders within 4 areas, as shown in Figure 1.

illustration not visible in this excerpt

Figure 1: Stakeholder mapping - the power interest matrix

The importance or power and the level of interest of the stakeholders have to be part of the research. It needs to be deduced from the actions and reaction of the stakeholders, following certain recall incidents, what their point in this matrix tend to be. The literature could not give sufficient or broad insight into this topic. Therefore even the obvious stakeholders like the company, the physician, the patient and his/her relatives, the state, or the regulatory institution could not be placed into this matrix based on any available scientific publication available. Still this matrix will help to understand, organize and interpret the qualitative materials gathered during the research.

2.10 Corporate communication models

There is a vast amount of literature about corporate communication. For this literature review and this research three theories where selected, because they tend to be respected and cited a lot and also show interesting and more importantly different aspects of corporate communication. All three of them seem to become important within this research. This chapter is based upon Hübner2007, which gave a fantastic introduction and insight on the literature of this topic.

Argenti2003 shows the American view on corporate communication, whereas Bruhn2003 and Cornellision2004 show the European view. These three models should provide a sufficient basis to analysis the communication strategies from different viewpoints in chapter 4.

2.10.1 Argenti’s functional definition of corporate communication

Argenti’s main aspects on corporate communication are:

- Companies need to communicate strategically
- They should communicate in loops as follows:
- The company communicates with messages
- to its constituencies (stakeholders or target groups)
- who then respond to
- the organization
- The following steps need to be taken:
- analyse the companies objectives, resources, reputation
- analyse constituencies (stakeholders or target groups)
- decide how and by what approach messages are delivered
- measure constituency response to measure success

(Argenti2003, p. 21-31)

This model is important because this research aims to derive strategies from public communication of the company and the other stakeholders (constituencies) in RQ1 and wants to evaluate their success in RQ2. Both aspects are part of Argenti’s model.

Key to all message delivery through different channels is the concept of visual identity (Argenti2003, p. 58). This visual identity promotes an image of the company which the target group holds. So, if identity and image are aligned, a solid reputation can be maintained or build. Reputation is different from image, because it is build and perceived as build up over time. Thus it becomes an investment measured like an intangible asset (Lev2001).

2.10.2 Bruhn’s concept of integrated corporate communication

Bruhn has also a functional perspective on corporate communication but criticises the lack of focus on how other models structure corporate communication. Bruhn’s model contains three aspects: (1) content, (2) formal aspects, (3) timing. Integration means that all three aspects need to be integrated in a way that even if different target groups are addressed, the instruments need to follow a consistent set of rules. This supports the overall goal: All target groups should be addressed with an aligned content, corporate design, and timing to ensure that the different communication activities support each other. All these alignments should be documented with written goals, target groups and core instruments, as well as all rules that apply for these instruments. This provides concrete instructions for communication managers (Bruhn2003).

Bruhn’s model is important for this study because it may allow deriving of the rules of such a virtual document from the corporate communication that will be found in this research.

2.10.3 Cornelissen’s approach to the organisation of corporate communication

Cornelissen’s definition of corporate communication:

Corporate communication is a management function that offers a framework and vocabulary for the effective coordination of all means of communications with the overall purpose of establishing and maintaining favourable reputations with stakeholder groups upon which the organization is dependent (Cornelissen2004, p. 23)

The reputation is based on three elements (Cornelissen2004, p. 83ff): (1) it is a perceptual construct, (2) it is formed (multiple times) by several stakeholder groups, (3) it includes an evaluation by the stakeholders so that it can differ from the intended corporate identify.

Cornelissen’s definition is of special importance for this study because it introduces the stakeholder view into the functional corporate communication definition giving a more inclusive view (Cornelissen2004, p. 24). Additionally this definition highlights the importance of reputation that especially in crisis communication was found to be a very important factor for the outcome of crisis communication or activities around a crisis (Coombs2004).

If reputation could be measured in this research this would very much support the use of those three models or parts of them for the analysis and to answer the research questions.

2.10.4 Corporate communication in context

Jarzabkowski2003 developed a framework for interpreting and analysing strategy practice, which is also useful to analyse communication strategies. Central to his framework, translated into the corporate communication domain, is that managers interact with stakeholders, history, culture, and engage in a form of shared endeavour in strategic activity (Jarzabkowski2003, 25). Three conclusions can be drawn from his work in conjunction with the other models: (1) Corporate communication is not only a vehicle to communicate between different parties it is embedded into a specific context which may support or not support the mediation between the parties. (2) Corporate communication can generate meaning from and impose meaning upon the context (Jarzabkowski2003, 26). (3) Corporate communication needs to be analysed with respect to actors, structures, and activities in order to show the whole interplay in corporate communication practice.

3 Research methodology

3.1 Research strategy and design

The research aims to find an answer to the question – Is there a favourable way in the content and utilization of communication channel before, during or after a product recall of branded products? Qualitative data will be analysed and theories of corporate and crisis communication will be used to help to understand and generalize the findings. This will lead to answers to the RQ1 – finding the communications strategies of the different companies operating in the implantable device business. Multiple case-studies are necessary with a longitudinal research approach for multiple companies. But the research also incorporates cross-sectional aspects, when responses of those companies are analysed against external stimulus. This approach allows analysing three important aspects that are key issues in qualitative research [Silverman2005, p107]: chronology, context, and comparisons. Chronologic observations allow looking for process changes and developments, contextual analysis allows checking for special circumstances that wouldn’t allow generalizations. And a comparison forces to divide data into sets, to make the data comparable.

The second objective is to finally find recommendations for communications strategies (RQ2). Thus the research needs an inductive or interpretive approach.

Due to the fact that the overall research method in the master thesis is qualitative research (i.e. text and source analysis), the research design is rather flexible. A more flexible research design is recommended when primarily qualitative methods are incorporated (Robson2002).

Product recalls, related communication and related business performance will be analysed and an underlying communication strategy of the company will be induced by interpretation. The induced communication strategy will then be related back to the business performance to build a theory – a thesis – that can of course be tested in further research. Creating a theory (incorporating existing theories of course) is necessary due to the open research questions. Further research should later collect and analyse more and newer data, to test the theory.

[...]