Litigation claims relating to venous thromboembolism in the NHS

Introduction

Venous thromboembolism (VTE) in hospitalised patients is associated with substantial morbidity and mortality. In 2004–2005, the House of Commons Health Committee reported an estimated annual preventable death rate of 25,000 people from VTE in the UK. ¹ The same Committee estimated the cost of VTE management at £604 million. This sum reflects the protracted morbidity associated with conditions such as deep vein thrombosis (DVT) and pulmonary emboli (PE) as well as their sequelae: post-thrombotic syndrome (PTS) and chromic pulmonary hypertension (PH). These costs are also a consequence of the high incidence of VTE. The incidence of VTE has been estimated using epidemiological models, for example Cohen et al. ² estimated that the annual number of DVTs across six European Countries, including the UK, was close to half a million events (between 404,664 and 538,189).

Whilst these estimates confirm the substantial health burden that VTE carries, multiple clinical trials have demonstrated that VTE is often preventable with appropriate prophylaxis. Such studies have provided high level evidence for the guidelines on prevention of VTE in hospital patients published by the National Institute of Clinical Excellence. ³

Litigation against the NHS is handled by the National Health Service Litigation Authority (NHSLA) through its risk pooling schemes. ⁴ Within the remit of its mandate is the assessment of all legal claims against the NHS including claims for medical negligence. Each claim is assessed, and where the legal action is justified settlement will be sought. Where an application for damages is not justified, the NHSLA will vigorously defend the case. Currently the NHSLA estimates that claims related to VTE have cost the NHS in excess of £20 million. This estimate does not detail the pattern of litigation nor does it identify the specialty areas of clinical practice with the highest number of VTE claims and hence those areas most vulnerable to VTE litigation and its financial burden. The aims of this study are therefore to identify the surgical and medical specialities where claims were made and to assess the corresponding financial burden of the claims. To do this, the nature of the VTE event and the damages incurred in each case were determined as well as the changing pattern of litigation with time. This analysis may contribute to improved quality of care by raising the awareness of this preventable and morbid condition and to establish if direct learning points are possible.

Methodology

An application was made to the NHSLA for information on medical negligence claims made against the NHS for injury caused by VTE events. Anonymised details of claims over a five-year period between 1 April 2007 and 31 March 2012 were returned by the NHSLA. The NHSLA identified each claim from their databases by using the search terms “thrombosis” and/or “embolism”. A short summary of the salient features of the individual cases was provided, together with the year of the incident and the perceived location of the incident, e.g. accident and emergency (A&E), the ward or in the outpatient department. The specialty department involved in the case was given, e.g. orthopaedic surgery, gynaecology or general medicine. The total sum of damages paid to each claimant was also disclosed. The NHSLA has a policy to limit the number of identifiable data items supplied to ensure patient confidentiality. Therefore, the age and gender of each claimant was provided on an aggregate basis rather than individually, with defined age ranges. It was possible to obtain individualised data on ages and gender of the claimants in each case, but this would have been at the expense of details of the individual damages paid. The activity status of each case was classified as either “closed” or “outstanding” as on 28 March 2013. Closed cases were defined as the completed cases where damages either had been paid out or where no justification for the legal action against the NHS could be found. “Outstanding” cases are where the legal action is still ongoing or pending. A formal disclaimer was given by the NHSLA that the accuracy and consistency of the coding could not be guaranteed because all the data were taken from a database which is designed to be used for claims management rather than a risk management tool. The NHSLA also specify that the data is provided without interpretation.

All the data were forwarded on an Microsoft Excel spreadsheet. In addition, the NHSLA provided an estimate of the projected costs of the outstanding claims. These costs included the anticipated total damages to be paid to claimants as a single group, plus an estimate of the total costs of litigation. The data on the Excel sheet were subcategorised for each claimant into:

(a) the nature of the VTE event; (i) PE, (ii) DVT, (iii) death or (iv) unclassified;

Where there had been multiple VTE events, categories were assigned as follows:

(a) if the patient had suffered a DVT and PE, the VTE event was recorded as a PE;

Patients were divided into three main groups: (1) “Completed” cases where the case was finalised and damages received for a VTE injury, (2) “Closed” cases where the case had been closed without payment of damages because no justifiable action against the NHS had been found and (3) “Outstanding” cases where the case still had legal action or a judgement pending.

Graphical representations of data were constructed and statistical analysis was performed using Graphpad Prism 4 software. Total amounts with median values and the ranges of payments were used to describe the damages incurred. Inter-group differences (damages for DVT, PE, death and unclassified) were evaluated by the use of ANOVA and the nonparametric Kruskal–Wallis test. Bartlett's test of variance was used to test for differences in the variance of the damages between the different VTE groups. Intergroup differences were evaluated by the nonparametric Mann–Whitney U test. Categorical variables were presented as frequencies, and percentages were compared by nonparametric Wilcoxon Signed Rank test.

Results

A total of 189 claims for injury caused by VTE were made by patients between 1 April 2007 and 31 March 2012. A total of 104 cases relate to female patients and 85 cases to male patients. The ages of the claimants ranged from 17 to 75 + years old, with the modal age categories being the 45–54-year-old age group (45 patients), closely followed by the 35–44-year-old age group (44 patients).

There were 39 closed cases where no justified case was found against the NHS, 51 outstanding cases and 99 completed cases with payment of damages for VTE injury. A total of 63 (63.6%) actionable VTE incidents occurred across six different surgical directorates and there were two cases where the surgical department could not be identified. A total of 11 (11.1%) cases occurred within the umbrella of “general medicine” and 10 (10.1%) cases were attributed to other medical directorates including respiratory medicine, renal medicine and cardiology. A total of 15 (15.1%) cases were against A&E and 1 (1%) case was against a general practice. The flow chart (Figure 1) maps out the distribution of specialty areas where VTE injury occurred. OPEN IN VIEWER

The total number of completed plus closed cases was 138. These cases cost the NHS £9.57 million (£5.78 million in damages and £3.79 million in legal costs). The NHSLA estimate that the total costs (damages + legal costs) of the 51 “outstanding” claims will be £9.73 million. Therefore, the projected costs for VTE claims over the five years covered in this study will be in the order of £19.3 million. The total damages paid for DVT, PE, deaths and unclassified VTE events are shown in Table 1. The median payouts and the ranges of compensation paid are also shown. Claims ranged from an isolated failure to diagnose and treat a DVT (£3000 damages paid) to complex cases where the VTE event appeared to be a component of the complications, e.g. in one case £100,000 damages were paid where the claimant suffered division of the common iliac artery during laparotomy. The highest amount of damages was for “failure to provide prophylactic heparin during knee replacement operation”; £1million in compensation was paid, but no other details of the case were provided. The differences in the amounts paid out for each type of VTE event were not significant either by parametric or nonparametric testing with ANOVA and the Kruskal–Wallis test, respectively, but there was significant variance between those patients who had DVT, those who had PEs, those who died and those whose VTE event could not be classified, P < 0.0001.

OPEN IN VIEWER

Table 1.

Number of completed claims and their amounts.

	“Completed” claims	Total financial compensation (£)	Median payment (£)	Range (£)
DVT	34	1,521,186	18,250	3,000–300,893
PE	24	785,473	15,000	4,000–160,000
Deaths	14	1,316,063	30,000	3,000–370,000
Unclassified	27	2,156,769	25,000	3,000–1,000,000

The division of this financial burden is skewed towards surgery. A total of £3,257,394 was paid out by surgical specialties. Orthopaedic surgery carried the highest burden at £1,636,061 (median payment £22,500 (£7,500–£1 million)) with obstetrics & gynaecology carrying the second highest burden at £679,957 (median payment £20,000 (£5000–£270,000)). A&E, considered as an independent specialty for the purposes of this study, paid out £999,103 (median payment £12,500 (£3000–£300,893)). The medical specialties paid out a total of £1,532,996 of which general medicine paid £614,170 (median £15,000 (£3,000–£370,000)) and respiratory medicine had the second highest burden with a single payment of £360,000 for a death following diagnosis of a DVT. The full details of damages paid by each specialty are shown in Figures 2 to 4. OPEN IN VIEWER

OPEN IN VIEWER

Figure 3.

Both number and value of claims (log₁₀) in pounds sterling by individual medical speciality.

OPEN IN VIEWER

Figure 4.

Value and number of claims (log₁₀) in pounds sterling, in both A&E and general practice.

There were no significant differences in the amounts of the damages paid out for VTE injuries in each of the different surgical specialties (ANOVA and Kruskal–Wallis tests). Likewise, there were no significant differences in the amounts of damages paid out in each of the different medical specialities (ANOVA and Kruskal–Wallis tests). When the medical specialities were compared with the surgical specialities, there was no significant difference in the values of the damages paid out for VTE injury p = 0.47 (Mann–Whitney U test) and no evidence of significant differences in the variances p = 0.50 (F test).

The completed and outstanding groups were analysed with respect to the year in which the VTE injury occurred and the year in which the NHSLA was notified that a claim was pending. The result of this analysis is shown in Figure 5. The data show that the peak number of events occurred in 2008/2009 (42 cases) and the lowest recorded number of cases occurred in 2003/2004 (two cases). Since 2008/2009 the number of cases has tailed off to five events in 2011/2012. If this is compared with when the claims were notified to the NHSLA (Figure 6), it can be seen that the number of claims year on year since 2007/2008 has increased steadily to 2011/2012. These data do not include the closed claims or the number of VTE events that have not been identified to the NHSLA. OPEN IN VIEWER

OPEN IN VIEWER

Figure 6.

Number of claims by year according to the date of notification and whether they have been completed.

Discussion

The aims of this study were to identify those areas of surgical and medical practice where medical negligence applications were made against the NHS for injury caused by VTE and whether this knowledge could inspire continued quality improvement. In addition, the financial burden of the different VTE events, DVT, PE and deaths were assessed. The pattern of litigation was also reviewed. The data have shown that the surgical specialities have paid out more in damages (£3,257,394 in total) for injuries related to VTE than the medical specialties (£1,532,996 in total). Orthopaedic surgery paid the highest individual damages (£1 million) and the highest total compensation (£1,636,061). A&E paid nearly £1 million in damages (£999,103). General medicine paid out more damages than any other non-surgical specialty (£614,170). When analysis of the payments in each directorate was assessed no significant difference was found. This is reflected in the remarkably similar median values for damages paid within the surgical specialties (Figure 2). Although the median values for the different medical specialties are more variable (Figure 3), the greatest burden of claims was suffered by general medicine with median damages of £15,000 of a similar order to the median pay out by the surgical specialties.

Analysis of damages received for DVT, PE and death was not significantly different. This lack of significance remained even after removal of unclassified cases from the data. When analysed using ANOVA there is no significance between the groups, but the Bartlett's test for variance shows that there is significant variance in the range of damages paid between the groups (p < 0.0001). This variance may be due to the case mix; £3000 for failure to diagnose a DVT is different from £160,000 in damages for a breast cancer patient undergoing chemotherapy who “… Presented with breathlessness? PE but tests were inconclusive. Discharged with advice to return if symptoms returned. She did 3 days later and perfusion scan showed pulmonary embolus causing lung collapse”.

There were a number of potential limitations with the data. For example, accurate details of the nature of the claim could not be obtained from the data provided by the NHSLA. No details were available about what proportion of the damages in the cases were for the VTE event or indeed other aspects of the case such pain and suffering, loss of earnings and the cost of care. The complexity of an individual case and/or the impact that the VTE event may have had in generating further complications can only be inferred from the value of the damages. For example, one patient was awarded £270,000 for “negligent failure to recognise signs and symptoms of DVT”, whilst another patient received £4000 for “failure to treat DVT appropriately leading to PE”. The incomplete nature of some aspects of the data makes any firm conclusions about the true impact of the VTE on both the patient and its actual cost to the NHS unreliable. Furthermore ascertainment of the true number of VTE events within the dataset can be questioned; the NHSLA do not guarantee the quality of the data because the initial data are put into their data base by non-medical staff. Twenty-five percent of the cases (51 cases) in the data set were unclassified and some of these may not have been for a VTE event, e.g. £10,000 for a “blood clot” that may have been a haematoma because “the claimant alleges that a nurse was lifting her out of a chair into bed when the nurse banged her leg resulting in a blood clot”. The limitations of NHSLA data are also acknowledged by other authors. For example, Szypula et al. ⁵ recognised that the clinical details could not be verified and that more than 10% of the cases in their data set, as assessed by the authors, were misclassified as “anaesthesia”. This raised the possibility that there were anaesthesia claims which were not included in their original data because of misclassification in the NHSLA database.

Despite the shortcomings of the data, one interpretation of the data suggests that the number of VTE events coming to litigation gradually increased from 2003/2004 to 2008/2009 and was then followed by a steady fall in the incidence after 2008/2009. For reasons given above (incomplete data, spurious cases, etc.), the accuracy of this finding is uncertain. It is possible that the incidence of negligent VTE injury has fallen because more patients admitted to acute care hospitals are risk assessed for VTE as part of continued quality improvement across the NHS. This is evident by the fact that in July 2013, 96% of the 1.2 million adults admitted to NHS funded acute care providers were risk assessed for VTE on their admission. This compares with July 2010 when less than 50% of the adult admissions were risk assessed for VTE. ⁶ The impact of this change in practice will lag behind litigation activity, but if the outstanding claims in the 2011/2012 category are eventually “closed” as opposed to “completed”, this would support the perceived downward trend in VTE injury and improved quality of care. An alternate interpretation of the data suggests that the number of medico-legal claims for VTE injury has increased when analysed by year of notification. Again although it has to been seen in the context of the limitations of the data set provided, there is anecdotal evidence to support this suggestion. Khan et al. ⁷ suggested that the increased number of claims may be attributed to the higher expectations of patients with regard to their treatment and a greater awareness of the legal requirements for a reasonable standard of care. With these factors, and others, a patient's need to understand why an event occurred, may account for why there is an increase in medico-legal activity. Furthermore, there is evidence for a global increase in clinical negligence claims handled by the NHSLA ⁸ since the early 1990s. However, an increase in the number of claims with respect to VTE events may be spurious. Patient expectations are an important motivator for change of practice; in some cases this may have contributed to the emergence of defensive practice, ⁹ but it is also a driver to change behaviours that benefit patient care. For example, Goldenberg et al. ¹⁰ have shown that the number of claims for injury from MRSA infection has decreased in line with the reduction of bloodstream infections with this organism. The impact of the NICE guidelines on VTE on clinical practice has also been seen. In addition there is effect of The Limitation Act 1980 (c. 58). ¹¹ Briefly, this statute sets out the time within which a claimant is required to bring legal action for a breach of duty of care. It is reasonable to infer that while some of the earliest cases are now time barred, some of the more recent VTE events will still be within the statutory period. As a consequence, “latent” cases will continue to accrue because of a legally justified lag-period for “notification” of the case to the NHSLA. Thus, statutory limitation may explain why one interpretation of the data suggests an increase in the number of medico-legal claims with respect to VTE injuries in contrast to another interpretation that indicates that claims for these injuries are becoming less frequent.

Where this study can have an impact on quality improvement is in raising the profile of NHS litigation for VTE. Although the data used in this study were acquired for the purpose of litigation, they also carry with them learning points that will contribute to quality improvement in patient care. Quality improvement should be encouraged locally through thorough data acquisition and analysis.

Although there are no reliable data on how many VTE events there are in the UK, the number of medico-legal claims notified to the NHSLA may reflect a trend in the incidence of events. If this is true, the evidence to suggest a fall in medico-legal claims, together with the accumulating trial data on the benefits of VTE prophylaxis suggests the possibility of a declining incidence of VTE events. With this, the costs of litigation against the NHS for failure to manage VTE in hospitalised patients might also diminish. Inevitably, this will have a positive impact on the surgical and medical specialities where claims for VTE have been made. We would, however, advocate local and national audit of the prevalence of VTE in the hospital setting to further improve patient safety.