Avon and Wiltshire Mental Health Partnership NHS Trust

8. Ms A complains that the Trust failed to appropriately assess her daughter, Miss R’s, level of risk on 3 and 4 March 2024. She adds that it also failed to assess and identify appropriate social support and take appropriate action to safeguard her after she disclosed ongoing thoughts of suicide and attempts to end her life.

9. She adds the Trust failed to adequately address her concerns in its complaint response.

10. Ms A also complains that a GP at the Practice prescribed an inappropriately large amount of antidepressant medication following a consultation on 1 March 2024. She adds the GP failed to explain the side effects of the medication, including that they may make her feel worse, and check whether Miss R had any thoughts of suicide or plans to end her life.

11. She says these failings led to her daughter taking her own life on 5 March 2024. She adds this has been devastating for her and has caused her own mental health to deteriorate significantly.

12. She would like both organisations to acknowledge what went wrong and improve their service, so these mistakes are not repeated. She would also like the Trust to compensate her financially for the impact its failings had on her.

13. Miss R was a woman in her mid-20s with no prior contact with mental health services. In early 2024, her mental health deteriorated quickly following the breakdown of her relationship with her boyfriend.

14. On 1 March, Miss R attended a GP appointment at the Practice. She scored highly on the PHQ-9 scale (a questionnaire which assesses depression), indicating severe depression. She told the GP she was feeling low, was finding things ‘too much’, but was able to go to work.

15. The GP prescribed sertraline, an antidepressant medication, and made an appointment for 2 weeks’ time with a mental health nurse who worked for the Practice.

16. The following day, Miss R presented at a Minor Injuries Unit following self-harm to her wrist that required medical intervention. On 3 March her father called 999 due to concerns that she may be at risk of self-harm and suicide. Miss R was referred to the Trust’s community mental health services.

17. The Trust contacted Miss R that day and documented she felt there was no point in her being alive, that she had a history of abuse, and would not agree to a safety plan because she wanted to ‘end it all’ and there was ‘nothing left to live for’. The outcome of this call was for the Trust’s Emergency Mental Health Triage team to attend later that day. The Trust advised her father to call 999 if he became concerned for her safety.

18. The Emergency Mental Health Triage team assessed Miss R later that evening. The assessment noted attempts to end her life, including cutting her wrist the day prior and that she had tried to drown herself in the bath. It also noted that she had written suicide notes to her family, that she was having ‘intrusive’ suicidal thoughts, and had thought about a lot of different ways to end her life. The outcome of the assessment was to discuss Miss R’s case in the next multidisciplinary team (MDT) meeting.

19. The following morning, Miss R’s father called 999 as he had awoken to the sound of his daughter trying to choke herself. The emergency services attempted to contact the Trust’s Intensive Support Team, but it was unable to respond. It advised it would contact Miss R later that day for an assessment.

20. Miss R was assessed by the Trust’s Intensive Support Team, and her case was discussed at an MDT meeting later that afternoon. A plan was documented to complete a safety plan and discharge her back to the care of her GP. She was advised to call the team if she felt low. They documented that Miss R agreed with the plan.

21. The next day, Miss R ended her life. The Intensive Support Team were notified of her death by the police, and it sent duty of candour letter to her family on 7 March.

22. The Trust undertook a Patient Safety Review, which concluded Miss R did not meet the criteria for more ‘assertive interventions’ and that Miss R had been known to the service for less than 24 hours.

Risk assessment and support (the Trust)

26. NICE guideline NG225 (self-harm) recommends that practitioners should carry out a psychosocial assessment after an episode of self-harm (1.5). This should include (1.5.10):

• historic factors • changeable and current factors • future factors, including upcoming events and circumstances • protective or mitigating factors.

27. This psychosocial assessment should include a risk formulation, which is a collaborative process between the professional(s) and the person that summarises their current risks and difficulties (1.6.6). This formulation typically includes historical factors, more recent problems, and existing strengths and resources.

28. These guidelines also say that services should not use risk assessment tools that stratify risk as ‘low, medium, high’ to predict future suicide or repetition of self-harm (1.6.3). Instead, risk formulation should focus on the person’s needs and how to support their medium and long-term safety (1.6.5).

29. NICE guideline NG225 also says that healthcare professionals should develop a safety plan with the person. This should include establishing the means of self-harm or suicide, triggers and warning signs, coping strategies, and safeguarding the environment by removing or restricting access to lethal means (1.11.7). A copy of this plan should be given to the person (1.11.8).

30. The Trust has provided evidence that between the time NICE published its updated guidelines for self-harm in September 2022 and Miss R’s presentation at its services in March 2024, it had been undertaking a systemic and long-term project of improving risk assessment and safety planning across the Trust. These changes aimed to bring its risk assessments in line with NICE guideline NG225. This included no longer using high/medium/low ratings for risk assessments.

31. The Trust explained it would not be safe to implement such widespread policy changes without ensuring all staff had been trained, it had the required policies in place, and it was satisfied the changes could be implemented safely. We agree with the Trust’s assertion that such systemic changes would need to be approached carefully and safely.

32. The Trust has also explained the final phase of the project was due to go live on 2 April 2024, just one month after Miss R’s tragic death.

33. The police referred Miss R to the Trust’s Emergency Mental Health Triage team on 3 March. It assessed Miss R that day and documented a psychosocial risk assessment, in line in NICE guideline NG225. This included an assessment of historic factors, her current situation, risks and safety netting advice, and involved her father in the assessment.

34. Miss R was referred to the Intensive Support Team by the Emergency Mental Health Triage team on 4 March. Two clinicians from this team assessed her face-to-face later that day. The clinicians recorded a detailed assessment in Miss R’s notes which took a similar structure to the assessment undertaken by the Emergency Mental Health Triage team.

35. The psychosocial assessment included an assessment of Miss R’s current circumstances and historical factors. It also included her father’s point of view, though we note Ms A questions his ability to engage because she says he has an intellectual disability. It noted current risks and that the team had advised her to call them if she felt low in mood. Following the face-to-face assessment, the Trust completed a risk assessment tool that graded Miss R’s risk to herself as medium.

36. Miss R had several factors that put her at high risk of attempting/completing suicide. She was experiencing hopelessness, the loss of a relationship and had made multiple impulsive and violent attempts at suicide. Previous instances where she had attempted suicide appear to have occurred without any prior planning, using what was in her immediate environment. Given her level of impulsivity, it is unlikely she would have contacted mental health services before acting on her thoughts of suicide. The Trust did not document how or why it reached the conclusion that contacting them would be an effective safeguarding strategy, especially as all previous contacts/referrals had been via a third party, not Miss R.

37. The methods of ending her life that that Miss R had already engaged in and/or described did not require long-term planning. This meant there was an increased risk if she acted impulsively on her thoughts of and suicide. She had also written suicide notes, indicating her intention to end her life as opposed to self-harming as a coping strategy. It was also notable that her father expressed concerns about how changeable his daughter’s mood was and how ‘rapidly’ this happened.

38. Following the face-to-face assessment on 4 March, the Intensive Support Team told Miss R to contact them if she felt low. It then discussed her case with the multidisciplinary team and agreed a plan to complete a crisis and contingency plan and signpost her to community crisis services. This work was proposed to start on 7 March. The Trust documented that Miss R had agreed with this plan but there is no evidence that any written copy of a crisis or safety plan had been given to her, nor that any safety plan to address her immediate risk had been agreed other than to contact the mental health team.

39. There was little planning documented around keeping Miss R safe in the immediate term, between 4 and 7 March. She had been advised to contact the team if she felt low, but she was not given any other options for managing her rapidly changing mood, nor did this action adequately address the impulsive nature of her self-harming actions. Given how rapidly her mood had been changing, how impulsively she had acted on her thoughts of suicide, and the fact she had never reached out for help during her prior instances of self-harm, it was unlikely she would call the mental health team or any crisis service at the point she felt an impulse to end her life. Our Mental Health Nurse adviser said Miss R ‘was not help seeking at this point, therefore having a safety plan that [relied on] Miss R [seeking] help [was] not robust enough. This does not appear to have been considered by the Trust’.

40. We have found the Trust’s failure to agree an immediate-term safety plan that was appropriate to Miss R’s presentation, in particular her level of impulsivity and access to lethal means, falls short of NICE guideline NG225.

41. We recognise the Trust was planning to complete a crisis and contingency plan with Miss R over the coming weeks. It has also said a comprehensive safety plan can take a number of contacts to formulate and its focus was on immediate term harm at this time. We agree a fully comprehensive plan would take time to formulate, but this should not have prevented the Trust from taking appropriate steps to safeguard Miss R in the immediate term.

42. The only safety plan in place in the days prior to her death was for her to call the mental health team or a local crisis service for support. There was no documented consideration as to how this would address Miss R’s risk in the immediate term, especially in the context of her rapidly changing mood, impulsivity, and violent and escalating methods of self-harm. This plan was insufficient to meet Miss R’s needs in the immediate term.

43. There was very little consideration of how to manage Miss R’s risk from 4 to 7 March, and the Trust did not document safety planning around reducing her access to the means of suicide, as recommended by NICE guideline NG225. For example, Miss R used her motor vehicle to die by suicide and her access to this could have been made more difficult or removed by asking her to give the keys to a friend or family member.

44. We have found the Trust undertook a psychosocial assessment in line with NICE guideline NG225. However, it overlooked key aspects of her immediate term risk, and it did not then document how it had reached its conclusion that the agreed plan would manage the risks identified.

45. We acknowledge that grading a risk of self-harm or suicide as ‘medium’ is not recommended by NICE guideline NG225. That said, the Trust has provided comprehensive evidence it was addressing this as part of a long-term project of systemic service improvements to bring risk assessments in line with NICE guideline NG225. Based on this evidence, we have not found that grading this risk as medium was service failure.

46. There was also a lack of professional curiosity around Miss R’s level of risk and how the Trust could prevent the risk of harm. Professional curiosity can be defined as having a ‘healthy scepticism’ about the information being communicated rather than accepting it at face value. This curiosity enables opportunities to identify less obvious indicators of vulnerability or protective factors. This is particularly apparent in the Trust’s confidence that Miss R would not act on her thoughts of suicide because she had told them this.

47. Professional curiosity in this case, for example, could have led the Trust to consider whether Miss R’s statement that she did not intend to hurt herself was congruent with her actions over the preceding days, and whether her level of impulsivity and rapidly changing mood meant her ability to predict if and when this could change was limited.

48. The Trust also did not document an exploration of Miss R’s triggers or the violent and impulsive nature of her acts of self-harm. Miss R’s father also told the Trust how ‘rapidly’ things could change, but the Trust did not document any consideration of how to manage this risk in the immediate term.

49. We have found the Trust missed an opportunity to ensure the risk management plan included actions that were appropriately tailored and proportionate to Miss R’s individual level of risk, including triggers and access to lethal means. This plan did not need to be fully comprehensive; rather, it needed to be sufficient to manage her level of risk in the immediate term until a more comprehensive assessment and planning could take place. We have found this did not happen. We recognise, however, that these risks can never be entirely removed in a community setting, and that all NHS mental health service providers must balance risk management with positive risk taking.

50. We have found the lack of detail in the risk assessment and the failure to ensure the risk management plan appropriately addressed key risk factors amounts to service failure.

51. Although there were failings in the assessment and safety planning around Miss R’s level of risk, this does not necessarily mean different actions should or would have been taken. As a general principle, practitioners must engage with the least restrictive approach that is possible and, without a Mental Health Act assessment, all interventions would need to have been agreed informally with Miss R. Practitioners must also consider whether a more forceful approach could impact on the trust and rapport that is being built with the person.

52. We cannot be certain of how willing Miss R would have been to engage with other interventions, and whether she would have agreed to interventions aimed at restricting her access to the means of ending her life. If she had agreed, it is unlikely access to every possible and foreseeable means of suicide could have reasonably been removed or restricted. We also cannot know whether, if offered, she would have accepted or engaged with more intensive options for support.

53. The Trust would not have been able to compel Miss R to engage with these interventions without arranging for an assessment under the Mental Health Act 1983. This is the most restrictive approach possible because it involves depriving a person of their liberty. Because of this, it is only used as a last resort when the risk of harm is so extreme it outweighs the risk of harm from depriving the person of their freedom.

54. We also have the bear in mind that when reviewing what could have been different, we have the benefit of hindsight. We know the tragic outcome that occurred, and this is information the Trust did not know at the time it assessed Miss R.

55. Our view is that there are too many unknowns about what could have happened if the Trust’s risk assessment had fully considered Miss R’s risk and documented a plan that specifically aimed to address these in the immediate term. Whilst it is possible that more intensive or assertive actions could have been considered, we cannot know, even on the balance of probabilities, whether these would have prevented the tragic outcome in this case. The benefit of hindsight can make it appear self-evident that more intensive support should have been offered, but with the information available to the Trust at the time this would not necessarily have been the case.

56. We recognise, however, how distressing these events were for everyone who knew and cared for Miss R, particularly her family. We also recognise that Ms A has been left devastated by the loss of her daughter and that this has caused her own mental health to deteriorate. We cannot robustly conclude, however, that the Trust’s actions directly caused this distress. This is because we cannot know whether Miss A’s death could have been avoided, and so we cannot say the distress arising from her death could also have been avoided.

57. What we can say is that knowing the Trust’s assessment and management of her daughter’s risk the day before her death was not completed to an appropriately detailed standard has amplified her grief at losing her daughter. Ms A says her daughter was her best friend and she has been left with unanswered questions about whether her daughter’s death could have been avoided had the overall risk assessment and safety planning been in line with NICE guideline NG225.

Complaint handling (the Trust) 58. Our NHS Complaint Standards offer a framework to support NHS services providers to handle complaints and take learning to improve services. In line with these the Trust should have:

• given a fair and balanced account of what happened based on established facts • clearly referenced any guidelines or policies and used objective criteria to make its decision(s) • when appropriate, explained why things went wrong and identified suitable ways to put things right.

59. Ms A complained about the care provided to her daughter on 16 September. In this letter she raised the following concerns:

• the Trust’s communication with Miss R’s family, especially regarding safety planning, was not to a good standard • the risk assessment conducted by the Trust overlooked important information • Miss R’s father has an intellectual disability and felt unable to safeguard her, and Miss R did not view her father as a protective factor • she does not believe her daughter had mental capacity at the time.

60. The Trust issued a response on 23 October. This response detailed that Miss R’s father was included during safety planning when she was assessed at home. However, it added Miss R had asked he not be present for other parts of the assessment. It agreed that it should have asked if there was anyone else the team should contact to support Miss R and that it was updating its processes to include this.

61. It also said a full and comprehensive risk assessment had taken place by qualified professionals, and that risk management information had been handed over appropriately. It reiterated the risk was ‘explored fully’ and there was no reason to doubt Miss R’s mental capacity.

62. This response offered an incomplete account of what action was taken. Whilst it outlined how it had assessed Miss R, it failed to outline the action taken and why the Trust felt the safety plan was sufficient. It also failed to respond to Ms A’s question as to whether the impact of the medication Miss R had been prescribed was considered as part of the risk assessment.

63. The response also did not outline the Trust’s policies or national guidelines and did not say whether these were followed. The information provided also gives the impression a detailed safety plan was agreed, but there is no evidence of a detailed plan in Miss R’s medical records. Instead, an agreement to ring the mental health team or for her father to call 999 is the only evidence of a plan.

64. We have found the Trust’s first complaint response provided an incomplete account of the care provided to Miss R and failed to include relevant policies and/or guidance to explain how it had reached the conclusions made.

65. Ms A wrote to the Trust again on 5 December. She asked for more information about the response issued by the Trust. This mostly focused on details within its explanation of how it considered her daughter’s risk and disagreement with the decisions made as a result of the information it had at the time.

66. On 28 January 2025 the Trust issued its final response to Ms A’s complaint. This response clarified that some small discrepancies in the Patient Safety Review were made because of how soon it was completed after the events. It also confirmed that the learning from the Patient Safety Review had been shared with all services run by the Trust.

67. The Trust also explained which professionals had been engaged to undertake the review, reiterated that a full risk assessment had been undertaken, and ‘safety plans’ had been in place. This response mostly reiterated the Trust’s position from the first response and provided some additional information in response to specific questions asked by Ms A.

68. We have found that whilst the Trust was clearly attempting to address Ms A’s concerns, the responses included incomplete information and did not include any guidelines or standards. The responses both stated there was a safety plan in place and gave the impression more had been done to preserve Miss R’s safety than documented in her medical records.

69. We can see that the Trust was open to identifying areas where it could improve its service because it demonstrated this in the Patient Safety Review and its first complaint response. The Trust made some recommendations for improving services following the safety review and decided to share information with Ms A about these changes.

70. We have found the Trust was attempting to provide a good standard of complaint handling. It also demonstrated its openness to taking learning from what happened. That said, because the Trust had overlooked key issues in the risk assessment and safety planning, the information provided was incomplete. It also did not reference the standards to which it was working towards. This fell short of our NHS Complaint Standards and amounts to service failure.

Prescription and side effects (the Practice) 71. NICE guideline NG222 (depression in adults) says that clinicians should consider using a clinically validated scale to assess people presenting with suspected depression. This should be used to inform and evaluate treatment options (1.2.4). A clinically validated scale is a questionnaire that has been tested to make sure it is measuring what it is supposed to measure.

72. These guidelines also say that clinicians should conduct a comprehensive assessment that does not rely on symptom count alone. This should include the severity of symptoms, previous history, and the duration and course of the illness. It should also include an assessment of the impact of symptoms on the person’s functioning and/or disability (1.2.6).

73. The guidelines are clear that clinicians must always ask people with depression directly about suicidal thoughts and intentions (1.2.8). Additionally, they should advise that the initial stages of pharmacological treatment can increase agitation, anxiety, and thoughts of suicide, and advise them on how to seek help promptly if this happens (1.2.10).

74. The GMC’s Good Medical Practice guidelines say that doctors must record their work clearly, legibly and accurately. This should include any relevant clinical findings, decisions, and actions agreed upon. It should also include the information given to the person, information about any drugs prescribed, and any other investigations or treatment.

75. On 1 March, Miss R attended an appointment with a GP at the Practice. Prior to the appointment, Miss R had completed a clinically validated questionnaire (the PHQ-9) to assess her symptoms and their severity. Miss R scored highly on this questionnaire, indicating severe depression. The scale included a question about self-harm and suicide, to which Miss R indicated a low risk. This was in line with NICE guideline NG222.

76. The GP’s documented assessment of Miss R’s presentation contained several factors, including what had triggered her low mood, her functioning, and that she was struggling socially. This represented a comprehensive assessment, in line with NICE guideline NG222.

77. The GP did not document they asked Miss R directly about self-harm or suicide. Although she had indicated a low risk on the questionnaire, the question was limited because it does not explore specific thoughts or plans. Directly asking a person about suicidal thoughts and intentions is more reliable, and it enables a more detailed assessment of the risk. Patients may also feel more able to volunteer information if the clinician has built a rapport with them.

78. The GP has stated that although they did not document asking Miss R this question, they would have done this as part of their usual practice. The GMC’s Good Medical Practice guidelines state doctors must accurately document their work. It can be difficult for any clinician to recall what happened in an appointment that occurred some time ago, especially for doctors who will likely see a large volume of patients. In the absence of a contemporaneous record of what happened during the appointment, we cannot conclude this conversation happened. This amounts service failure.

79. The GP also did not document any discussion around the side effects of the medication they had prescribed. This was important because this type of medication can increase thoughts of self-harm and suicide in the initial weeks of treatment. The GP stated it is their usual practice to discuss this with patients; however, for the reasons outlined above, we cannot robustly conclude this happened. This amounts to service failure.

80. The GP booked a follow-up appointment with a Mental Health Nurse at the Practice for two weeks’ time, which was in line with NICE guideline NG222. They did not document any further safety netting or advice, such as how to seek help if Miss R felt suicidal or that she could not keep herself safe. This was not in line with NICE guideline NG222, which recommends giving information about sources of crisis support.

81. We have not found the identified failings directly led to Miss R taking her own life. She was seen by community mental health services following this appointment, and her risk was assessed by more specialist clinicians. What we can say, however, is that knowing the GP did not document an exploration of Miss R’s risk of suicide, an explanation of how the prescribed medication could exacerbate this, and provide signposting to crisis services, will cause Ms A distress.

82. With regards to whether it was appropriate to prescribe two months’ worth (56 tablets) of sertraline, there are no national guidelines as to how much medication should be prescribed. Whilst it can be safer to prescribe lower amounts of medication, there are other factors a doctor may consider. This can include the risks associated with that medication, whether the patient indicates there is a risk of overdose, and the financial burden that can present with multiple short-term prescriptions.

83. According to the British National Formulary, sertraline is a drug that presents a low risk when an overdose occurs. It is very rare for an overdose to result in severe or fatal outcomes. An overdose of sertraline more commonly results in symptoms such as nausea, vomiting, tremor, and drowsiness. In very rare cases, it can cause a condition called serotonin syndrome, which can be fatal without treatment.

84. Whilst the quantity of tablets was not inappropriate in and of itself, the GP failed to document any explanation of the side-effects, nor was any safety netting advice provided in relation to the prescription. The number of tablets prescribed does not amount to service failure; however, the GP should have done more to explain the potential side-effects, in line with the GMC’s Good Medical Practice guidelines and NICE guideline NG222.

85. Miss R did not take an overdose, and she was seen by the Trust’s community mental health services after this appointment took place. The Trust was aware of the medication prescribed and documented no concerns about this, telling Miss R it was her choice whether to continue taking it.

86. We cannot know, even on the balance of probabilities, whether this prescription contributed to Miss R’s increased agitation and suicidal thoughts and actions. The number of tablets prescribed is not directly linked to this sad outcome because Miss R did not take an overdose. What is notable, however, is that there appears to have been a lack of consideration of this risk by the GP. This will understandably cause Miss R’s mother distress.

87. The Practice failed to meaningfully acknowledge the failure to document key aspects of Miss R’s assessment in its response to Ms A’s complaint. Instead, it relied on what the GP said they would usually do but did not comment on whether this actually happened. At the very least, it should have acknowledged this failure in documentation was not acceptable and apologised for the distress this may have caused. This did not happen, and we have found the Practice has not taken sufficient action to put right the impact of what went wrong.Recommendations 88. We make recommendations in line with our Principles for Remedy which say public bodies should acknowledge failures, apologise, make amends, and use the opportunity to improve their services.

89. Our Principles say we aim to ensure the public body puts the complainant back in the position they would have been in had nothing gone wrong. If that is not possible, the public body should compensate them appropriately.

90. In line with our Principles we recommend the Trust:

• write to Ms A to acknowledge the failings identified and apologise for the impact these had. This should happen within four weeks of the date of our final report.

• develop an action plan to ensure its risk assessment and safety planning is tailored to patients’ individual presentation and needs. This should include exploring a patient’s access to lethal means, and how any proposed safety actions will address the patient’s current level of risk. The action plan should include what action will be taken, who is responsible for actioning these, how these will address the failings identified, and the target date for completion. This should happen within 12 weeks of the date of our final report.

• pay Ms A £1,200 in recognition of the distress caused by these failings. This should happen with four weeks of the date of our final report.

91. We also recommend the Practice:

• write to Ms A to acknowledge the failings identified and apologise for the impact these had. This should happen within four weeks of the date of our final report.

• develop an action plan to ensure all staff at the Practice are familiar with the requirements set out in NICE guideline NG222 and that all discussions around suicide are robustly documented. The action plan should include what action will be taken, who is responsible for actioning these, how these will address the failings identified, and the target date for completion. This should happen within eight weeks of our final report.

1. We have found that Avon and Wiltshire Mental Health Partnership Trust (the Trust) undertook psychosocial assessments of Miss R’s needs on 3 and 4 March 2024, in line with NICE guideline NG225. However, it failed to document important considerations of her risk, and the documented safety plan did not appropriately detail how the actions proposed would mitigate the risks identified. This amounts to service failure.

2. We cannot know whether Miss R’s tragic death could have been avoided had failings not happened. This is because there are too many unknown factors, and it is possible that a more robust risk assessment and safety plan could have led to the same tragic outcome. Knowing the risk assessment and safety plan should have been more robust, and the uncertainty around whether this could have changed the outcome, will cause significant distress to Ms A, who says these failings have amplified her grief at losing her daughter.

3. We have also found that whilst the Trust attempted to provide a fair and balanced response to Ms A’s complaint, the information included was incomplete and it failed to reference the standards it was relying on. This was service failure and compounded Ms A’s distress following the loss of her daughter.

4. We have found the number of tablets prescribed by a GP at the GP Practice in South Gloucestershire (the Practice) was acceptable and they arranged for a follow-up appointment to review this medication in line with NICE guideline NG222.

5. We have also found the GP failed to ask Miss R directly about her thoughts of suicide. They also did not advise her that the medication could affect her mood, agitation, and thoughts of suicide, and did not provide sufficient safety netting advice. This amounts to service failure.

6. We have not found this led to Miss R’s death because she did not take an overdose. We also cannot know whether the medication contributed to increased agitation and thoughts of suicide. That said, knowing these actions were not documented as being completed by the GP will cause Ms A distress.

7. We partly uphold this complaint and have made recommendations at the end of this report.