A Study of Medical Negligence Claiming in Scotland

The study examined the current system for medical negligence claiming in Scotland. The study was commissioned to inform the work of the No-fault Compensation Review Group which was established in 2009 to consider the potential benefits to patients of the introduction of a no-fault compensation scheme for medical negligence claims in Scotland.



A.1 The 'No-Fault' Working Group has recommended that there should be a move towards a 'No-Fault' System (NFS) broadly equivalent to the system which operates in Sweden. This appendix examines the possible expenditure implications of the Working Group's proposals based on data provided by CLO.

A.2 Any discussion of the expenditure implications of the proposed No-Fault Scheme has to be based on assumptions about how the proposed scheme will operate and how potential claimants will respond to it. Different people will have genuine differences of opinion as to what are reasonable assumptions about how different potential claimants will respond to NFS. In this report we will use two sets of assumptions to generate an upper and a lower estimate.

A.3 We would stress that the estimates we produce below are not predictions about what the scheme will cost in the future but are estimates of what public expenditure would have been in a typical year over the recent past had the proposed NFS been in existence. This allows us to use the information which is available on actual claims under the negligence based system to make judgements as to what would have happened had a No-Fault System been in place.

Existing system baseline estimate

A.4 The Working Group has proposed that consideration be given by the Scottish Government to a 'no-fault' scheme for medical injury along the lines of the Swedish model. Nevertheless, it has proposed that compensation be based on need rather than a tariff. We understand this to imply that compensation under the no-fault scheme will be broadly of the same magnitude as available in successful claims under the litigation system. It has also been proposed that claimants who fail under the no-fault scheme should retain the right to litigate and that claimants who fail in litigation should have a residual right to claim under the no-fault scheme.

A.5 Our approach to estimating the cost implications of introducing a no-fault scheme follows our earlier approach of using the data supplied to the research team by CLO. We use the data on closed claims produced in February 2010 to provide the costs of operating the current system and then estimate the likely cost of dealing with claims under the proposed system. This leads to an estimate of what expenditure would have been had the no-fault scheme been operating in recent years. This means that the expenditure patterns under the two systems can be based on realistic patterns of claims.

A.6 The Working Group has recommended that the proposed scheme be extended to all health care professionals in Scotland and not just those employed by NHS Scotland. However, the estimates presented in this appendix refer only to health care professionals employed by NHS Scotland. This allows us to make a direct comparison of expenditure with that of the current system.

A.7 We take the costs of the present system to be the costs associated with what CLO calls closed claims. By costs we mean the cost in terms of public expenditure. This expenditure includes the compensation awards paid to successful claimants, the payment of adverse costs and outlays to successful (and in some cases unsuccessful) claimants, chargeable costs of CLO staff time and CLO outlays and expenses. Figure A1 illustrates the broad categories of expenditure for the years from 2004 to 2009.

A.8 It should be noted that although the expenditures associated with settled claims have been rising over the period they have not risen as much as award levels and consequently have fallen as a percentage of award levels more or less continuously since 2004 (from 41% to 25%).

Figure A1

Figure A1

A.9 We have used claims closed in each year as the basis for calculating the expenditure involved in the present system. However, there is a degree of arbitrariness on which year a case closes, particularly where a claim is of very high value and takes a long time to settle. In order to account for this we average the figures illustrated over the period 2004 to 2009 to arrive at, as it were, what expenditure has been in a 'typical' year. This figure is £18,057,455 and is treated as our baseline estimate for the current system.

A.10 Readers are reminded that the estimates given in this appendix do not include the extension of the proposed scheme to private contractors to NHS (such as GPs, dentists and pharmacists).

Generating estimates of expenditure under NFS

A.11 In order to estimate the cost implications of the proposed scheme assumptions have to be made concerning a number of factors:

  1. What proportion of the claims which settle under litigation will settle under a no-fault scheme and what will be the average award?
  2. What will be the cost of reaching decisions under the new system?
  3. What proportion of claims that fail to settle under the litigation system will have an award made under a no-fault system?
  4. What level of awards will be made in those claims which settle under a no claims system but did not settle under the present system of litigation?
  5. How many additional claims will be made under a no-fault system compared to the litigation system?
  6. What proportion of these additional claims will receive an award?
  7. What will be the level of award for those additional claims that are successful?


A.12 To answer these questions we need to make some assumptions. We now turn to these assumptions and their justification.

Proportion of claims which settle under litigation settling under a no-fault scheme

Assumption 1a

A.13 As mentioned in paragraph A.4, the Working Group has suggested that the choice of whether to seek compensation under NFS or under the negligence system should be left to the claimant. Under the proposed NFS, fault will still require to be shown but negligence will not. Because of the difficulties, cost and risks of proving negligence we assume that all claims will begin in NFS. Although the Working Group concluded that a claimant who is dissatisfied with an award or who is denied an award under NFS should have the right to claim for negligence, we take the view that it will not be rational for a claimant to do so since the burden of proof, costs and risks are greater and time taken to reach an outcome is much longer. We thus assume for the purposes of our estimates that there is no leakage from the NFS to the negligence system.

A.14 The Working Group also recommended that awards under NFS should not be based on a tariff and should be no lower than could be attained under the negligence system. This implies that awards should not be reduced to reflect the reduced risks and delays in settlement under NFS as compared to the present system. We believe that this further bolsters the case for assuming that there will be no leakage from NFS back into the court system.

A.15 The implication of the discussion in the preceding paragraphs is that all cases which are settled under the present system will settle under NFS

The assumed average level of award under NFS for claims which already settle under the current system

A16 A judgement needs to be made as to what the average level of award for those claims which settled under the present negligence system would be had the NFS been in place. This requires an interpretation of how the Working Group's recommendations will work in practise. The Working Group took the view that awards should be based on need and that "therefore points to levels of compensation equivalent to those that would be awarded by a civil court." However, we have no data on what a court would award in a civil action arising from a medical injury which required that fault be proved but no proof of negligence was required because no such action exists in Scots Civil Law at present. What the adjudicator under NFS has to do is, in the light of what is likely to be different evidence presented by the claimant and by the defender (and assuming fault to be proven) arrive at an award. Under the Working Group's recommendations the claimant has three potential courses of action: a) to accept the award; b) to sue for negligence; c) to appeal against the adjudicator's award. In deciding the magnitude of the award the adjudicator is likely to take account of options b) and c).

A.17 Conceptually the adjudicator is in the same position as the CLO considering making an offer under the negligence system in the shadow of litigation. We have concluded in A.13 above that those who settled under negligence would not choose to go down the negligence route when the option of a similar award was available under NFS. Consequently, we take the view that on average an award equal to the award under negligence would be sufficient to dissuade the claimant from suing under negligence. What level of award would dissuade the claimant from mounting an appeal to the Sheriff Court or Court of Session as proposed by the Working Group as an appeal mechanism? The attractiveness of an appeal to the claimant will depend very much on the detail of the legislation implementing NFS. It will also depend on the claimant's attitude to risk. In making a decision on whether to accept the defendant's offer (or proceed to a court hearing) the claimant would be mindful of the risk that the case may not be successful at court and that he or she may have to pay legal (and other) costs if unsuccessful. Experimental evidence shows that people are, in general, risk averse. That is they will accept a sure thing (i.e. an offer) which is lower than the uncertain (probability-adjusted) expected value of the court award. This risk implies that the pursuer would, on average, settle for less compensation than would be awarded by a judge should the case ultimately be successful in court. As there would be no need to prove negligence but only fault (if the adjudicator rejected the claim) it is likely that a claimant would judge an appeal under NFS to be less risky. This could raise the minimum (risk-adjusted award) likely to dissuade a claimant from appealing against the award and consequently raise the award offer made by the adjudicator. What this award would be on average will only be possible to estimate after awards have been made, accepted or rejected by claimants and decisions made on appeals. It is possible that they could, on average, be the same as under negligence depending on how risk averse claimants are. On the other hand levels of risk aversion among claimants may be such that they will require a higher award to dissuade them from appealing. How much higher they would be is difficult to judge. For every ten percentage points that awards are higher than the awards under negligence expenditure will increase by £1.3M

Assumption 1b

A.18 Our lower estimate of the average award made under NFS for claims that settled under the existing system is the average award under the existing system for this group of claims.

Assumption 1c

A.19 Our upper estimate of the average award made under NFS for claims that settled under the existing system is that they will be 20% higher than under the existing system

Costs of setting claims under NFS

Assumption 2

A.20 The description of the Swedish system received by the WG makes no mention of any involvement of lawyers on either side when claims are being considered. The administrative cost is stated to be £2,268 per claim. Fenn et al (2004) used an administrative cost estimate of £2,000 per claim supplied by NHS for a Swedish no-fault system for England. Fenn et al also assume a lower level of award under a no-fault system. On the other hand the WG's report says that causation will still be an issue. This seems to imply there will still be an adversarial element in determining causation. For this reason we make the assumption that the costs of settling claims will on average be higher than in the Swedish system at £4,000. This is around the average expenditure on unsettled claims in recent years. We apply this cost to all claims under NFS. Although under the current system a large proportion of legal claims are settled out of court, legal costs would still be incurred. The calculations assume that payments under a no-fault scheme will be of the same magnitude as those under the current litigation scheme.

A.21 The consequence of this assumption on the costs of dealing with claims that would settle is shown in Figure A2. This illustrates the very large savings which could result from a switch to 'No Fault'. Clearly to the extent that the actual expenditure in a 'No-Fault' system deviates from this sum the savings will change.

A.22 This suggests that in a 'typical year' expenditure on these 'settled' cases under the proposed system would have been £710,667.

Figure A2

Figure A2

A.23 Assumptions 1 and 2 apply to all of the estimates which we make of expenditure under NFS.

Claims which did not settle under present system

Assumption 3

A.24 We now turn to what we can assume will happen to claims that did not settle under the present system if they took place under NFS. Under the Swedish system around 45% of claims succeed whilst under the current Scottish system around one third of claims succeed. If the success rate under the proposed 'no-fault' system were to be the same as in Sweden 25% of claims which do not settle under the present system would have to receive awards. However, the number of claims per capita in the Swedish system is many times greater than under the present Scottish system. There are a number of factors (e.g. claiming culture, litigiousness, acceptance of medical dominance) which may give rise to this difference but it is beyond the scope of the present project to explain differences across countries in making claims. Nevertheless, we believe it is reasonable to assume that among those claims which are made under the present system but which do not settle there is likely to be a higher proportion which would receive an award under the proposed NFS than the overall success rate in the Swedish system.

A.25 We examined the implications of two possible success rates for such claims under NFS.

Assumption 3a

A.26 The first is that 40% of unsettled claims under negligence will succeed under NFS. This would have the effect of raising the success rate of those claims which were made under the present Scottish system to 61% for a 'typical' year under NFS. This implies an increase in the success rate for the claims that are made under the present system of just over 73% (i.e. an increase from 35% to 61%).

Assumption 3b

A.27 The alternative assumption which we make is that the success rate for these claims under NFS should be treated as 60%. This would raise the success rate for those claims which were made under the present system to 74% for a typical year under NFS which implies it is 110% higher than under the present system (i.e. an increase from 35% to 74%).

A.28 Both assumptions 3a and 3b result in a much higher success rate than in the Swedish system (45%). This is because we assume that the claims which do exist under the current system will have more 'merit' than those in the Swedish system because they are derived from a system which imposes both a negligence and a causation test. It is unlikely that a high proportion of claims made under the present system would fail the causation test although some will.

Level of award for claims which did not settle under present system

Assumption 4

A.29 The empirical research reported in Chapter 3, indicates that, other things being equal, under the present system higher valued claims were more likely to settle than lower valued claims. Further analysis of the CLO data makes clear that the average value of awards in any year is highly skewed, being dominated by a few very high value awards (over £1M) but also including a small number of high value awards (between £100,000 and £1M). This suggests that the average award under the present system is unlikely to be a reasonable estimate of the average award for claims which do not settle under the present system but are likely to settle under NFS. It is also the case that awards under the present system are very much lower than claim values. Consequently the claim value of this category of claim is not a good estimate of what the value of award in this category of claim is likely to be under NFS. Having examined in detail the CLO data we believe that reasonable estimate of the average award in this category is likely to be £20,000.

Additional claims under NFS

Assumption 5

A.30 Under the proposed NFS for Scotland the number of claims for compensation is likely to rise. The requirement to show only causation and not negligence is likely to lower the barriers to succeeding and the costs of making a claim. Fenn et al (2004) conclude, on the basis of a large scale survey, that the introduction of a No-Fault system in England would be likely to increase the number of claims by 80%. There is no comparable survey data for Scotland. We take the view that such an increase is likely to be at the upper end of reasonable estimates. Consequently we make our estimates of the number of additional claims under a Scottish NFS on the basis of two possible assumptions.

Assumption 5a

A.31 The first is that the increase is 20% of the existing claims in a typical year under the present system.

Assumption 5b

A.32 The alternative assumption which we make is that the increase is 80% as predicted for England.

Success rate under NFS

Assumption 6

A.33 However, it should not be expected that all of these additional claims will succeed under NFS. As mentioned in paragraph 4.21 only 45% of claims under the Swedish NFS succeed. What is a reasonable estimate of the success rate under NFS in Scotland of the additional claims that will be made compared to the present system? We have no real empirical basis for such an estimate. Consequently, for the purpose of generating estimates we make two alternative assumptions.

Assumption 6a

A.34 Our lower assumption is that the success rate is 20%.

Assumption 6b

A.35 Our higher assumption is that the success rate is 60%.

Average award for additional claims arising under NFS

Assumption 7

A.36 The final assumption that is needed relates to the average award that would be made to the additional claims arising under a Scottish NFS. These additional claims arise because the implied cost of making a claim fall under a NFS system compared to a negligence-based system. It seems reasonable to assume therefore that these claims (which were discouraged under the negligence system) are likely to be of lower value than those that were made. Thus based on similar considerations to those discussed in paragraph A.26 the average award made to these additional cases will be very much below the average award made under the present system. However it is likely that this group will have an even lower value than those claims that are made but fail under the present system. We, therefore, use an average award value of £16,000.

A.37 It could be argued that the larger the number of additional claims the lower the average award would be. This is plausible but not inevitable. It should be noted that Assumption 7 is at variance with that made by Fenn et al (2004) in their study of England & Wales where they assume that the average award under an NFS will be 75% of the average award under negligence. Fenn et al (2004) did not have data on the size distribution of awards. The CLO data provides that for Scotland and justifies a very much lower average award for these additional claims than under the present system.

Estimates of expenditure

A.38 The preceding paragraphs have produced eleven plausible assumptions which could be combined in various ways to generate alternative estimates of expenditure under the proposed NFS. However, we believe that greater clarity will be gained if we combine the assumptions, all of which have some plausibility, in such a way as to produce a lower and an upper estimate of what expenditure would have been in a typical year under the proposed NFS system.

A.39 The lower estimate is generated by combining Assumptions 1a,1b, 2, 3a, 4, 5a, 6a, and 7. The upper estimate combines Assumptions 1a, 1c, 2, 3b, 4, 5b, 6b and 7. Any other combinations of these assumptions will generate estimated expenditures which lie between these two combinations. Table A1 overleaf tabulates the expenditure under the existing system along with the lower and upper estimates.

A.40 Expenditure in a typical year under the present system is estimated to be £18,057,455. The lower estimate for NFS is £18,357,455 and the upper estimate is £27,014,275. Both the upper and lower estimates involve very much lower cost of determining awards and of course obviate legal and other costs for both sides. The lower estimate raises the number of successful claims by 85% compared to the current and successful claims represent 54% of all claims. The upper estimate results in almost three and one half times more successful claims with an overall success rate of 68%. It should be noted that under the Swedish system 45% of claims are successful.

A.41 Estimates produced are not predictions of what a no-fault scheme will cost in the future but are estimates of what public expenditure would have been in a typical year over the recent past for cases handled by the Central Legal Office had the proposed no-fault scheme been in existence.

A.42 It is also the case that under the current system a large proportion of legal claims are settled out of court - no allowances are built into the figures to reflect this as it has been assumed that payments under a no-fault scheme would be of the same magnitude as successful claims under the current litigation system.


Estimates of Expenditure

Awards on Claims settled under Present System Expenditure on Settled Claims Expenditure on Unsettled Claims Awards under NFS to previously unsettled claims Expenditure on 20% additional claims Expenditure on 80% additional claims Awards to additional claims Total Expenditure in Typical Year
Present system £13,017,629 £4,093,562 £946,265 £18,057,455
No Fault - lower £13,017,629 £710,667 £1,302,000 £2,604,000 £402,533 £322,027 £18,358,856
No Fault - upper £15,621,155 £710,667 £1,302,000 £3,906,000 £1,610,133 £3,864,320 £27,014,275


1. The estimates produced above are not predictions about what the scheme will cost in the future but are estimates of what public expenditure would have been in a typical year over the recent past for claims handled by the Central Legal Office had the proposed NFS been in existence.

2. Paragraphs A.11 to A.39 explain and justify the use of the assumptions used to calculate the estimated costs. The costs are based on a range and combination of assumptions involving consideration of the following factors:

  • The proportion of the claims which settle under litigation settling under a no-fault scheme;
  • Cost of settling claims under a No-fault system;
  • Proportion of claims that don't settle under the litigation system that would receive award under a no-fault system;
  • Level of awards for claims which settle did not settle under present system;
  • Additional claims under a no-fault system;
  • Success rate of additional claims; and
  • Average award for additional claims.


Email: Fiona Hodgkiss

Back to top