Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Intravenous fluid prescribing errors in children: Mixed methods analysis of critical incidents

  • Richard L. Conn ,

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Validation, Visualization, Writing – original draft, Writing – review & editing

    richardlconn@gmail.com

    Affiliation Centre for Medical Education, Queen’s University Belfast, Belfast, United Kingdom

  • Steven McVea,

    Roles Conceptualization, Formal analysis, Validation, Writing – review & editing

    Affiliation Neonatal Intensive Care Unit, Royal Jubilee Maternity Hospital, Belfast, United Kingdom

  • Angela Carrington,

    Roles Conceptualization, Data curation, Investigation, Methodology, Project administration, Resources, Validation, Writing – review & editing

    Affiliation Medicines Governance Team, Belfast Health and Social Care Trust, Belfast, United Kingdom

  • Tim Dornan

    Roles Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Validation, Writing – review & editing

    Affiliation Centre for Medical Education, Queen’s University Belfast, Belfast, United Kingdom

Abstract

Introduction

Recent National Institute for Health and Care Excellence (NICE) guidelines aim to improve intravenous (IV) fluid prescribing for children, but existing evidence about how and why fluid prescribing errors occur is limited. Studying this can lead to more effective implementation, through education and systems design.

Aims

  1. Identify types of IV fluid prescribing errors reported in practice
  2. Analyse factors that contribute to errors
  3. Provide guidance to educators and those responsible for designing systems

Methods

Mixed methods observational study which analysed critical incident reports relating to IV fluid prescribing errors in children aged 0–16, occurring between 2011 and 2015 in UK secondary care. We quantified characteristics and types of errors, then qualitatively analysed narrative descriptions, identifying underlying contributing factors.

Results

In the 40 incidents analysed, principal types of errors were incorrect rate of fluids, inappropriate choice of solution, and incorrect completion of prescription charts. Prescribers had to negotiate complex patients, interactions with other practitioners and teams, and challenging work environments; errors resulted from these inter-related contributing factors.

Conclusions

This study highlights the diverse range and complex nature of IV fluid prescribing errors reported in practice. While these findings have the inherent limitations of critical incident reports, they point to areas of potential improvement in education and systems design. Practising prescribing in context, inducting doctors within the many specialties who contribute to care of children, and educating them in joint working with nurses and pharmacists could help reduce errors.

Introduction

Intravenous (IV) fluid therapy is routine yet potentially lethal.[1] Whilst recent National Institute for Health and Care Excellence (NICE) guidelines[2] will help reduce risk, realising improvement needs a clearer understanding of error. Hyponatraemia has understandably dominated attention[314]; much less is known about other types of error or, importantly, their underlying causes. Finding out how and why IV fluid prescribing goes wrong could guide educators and help develop safer systems of care.

Prospective research involving children has been limited, and adult research is of limited applicability. Adult in-patients were affected by errors in calculating fluid rates, choosing types of fluid, and completing prescription charts.[15] Whilst these are self-evidently applicable to children, many specific aspects of prescribing differ, including methods of calculating rates of fluid administration, use of glucose-containing solutions, protocols, and charts. The little we know about paediatric prescribing comes mainly from small-scale audits, assessing particular types of errors. Errors in rate arose from miscalculation, use of incorrect formulae to calculate maintenance fluids,[6,11,14,16] and exceeding maximum allowed volumes.[7,17,18] Regarding fluid choice, even after 0.18% sodium chloride was withdrawn from use,[4,7,17] prescribers frequently prescribed hypotonic maintenance solutions (such as 0.45% sodium chloride),[18] even when hyponatraemia had developed.[47,9,13] They completed prescription charts incorrectly and omitted calculations and monitoring data.[4,13,14,17] Faced with this limited evidence base, NICE pragmatically recommended education to improve prescribers’ knowledge, and system changes such as standardising fluid prescription charts. But since imparting knowledge or introducing guidelines, alone, has little impact on doctors’ behaviour,[19] a more detailed analysis of the causes of errors could help target education more effectively, and advance NICE’s important work.

Our aim was to identify types of errors and explore contributing factors–how and why errors occur—to help make fluid therapy safer for children. Critical incidents–events reported by healthcare staff which cause actual or potential harm—have been established as a means to investigate errors. Reports provide categorical information, analysis of which enumerates the characteristics of errors, and narrative information, which helps identify causes. Researchers recently used this mixed methods approach to study patient safety issues and identify improvement opportunities in incident reports.[20,21] This article reports a mixed methods analysis of IV fluid prescribing incidents, categorising types of errors and identifying factors contributing to their occurrence.

Methods

Setting

This research was conducted within all five Health and Social Care Trusts in Northern Ireland, UK. Healthcare delivery in Northern Ireland is part of the National Health Service (NHS). Children are cared for in a broad range of settings: a large regional children’s hospital and neonatal unit; and several district general hospitals where inpatient paediatric wards provide medical and surgical care. Most hospitals also have maternity services and specialist neonatal units. Children may also receive care in non-paediatric settings, such as general emergency departments and specialist services that provide care for patients of all ages. In addition, children are typically moved to adult care once aged 14 or 15.

Staff voluntarily report critical incidents, using either an electronic database (Datix) available on hospital computers, or paper forms which are subsequently inputted to Datix. An example reporting form is available as S1 Form. All staff are encouraged to report incidents where harm occurred or there was a perceived risk of harm. Reports contain: patient demographics; where the incident occurred; what harm resulted; what type of incident it was; plus, a free text description of what happened and what subsequent action was taken.

Reports classified as medication incidents (including IV fluid incidents) are reviewed locally by medicines governance pharmacists (MGPs), whose role is similar to medication safety officers in other parts of the UK. They routinely review medication incidents and are trained in use of Datix. Categorical information within each reported incident is checked, including medication error type (prescribing/dispensing/administration/monitoring/other) and sub-type (wrong dose/wrong medicine etc.), drug(s) involved, and level of harm. Assignment of level of harm is guided by standardised risk matrices used in all Trusts (S1 Risk matrices).

Data extraction

MGPs within each trust extracted from Datix all medication incidents in patients aged 0–16, occurring between July 2011 and July 2015. Recommended age limits for paediatric care and clinical guidelines made this age range a logical choice. Moreover, adolescents have experienced morbidity related to inappropriate IV fluid therapy,[22] reinforcing the argument that they should be treated similarly to other children. All reports from incidents occurring in children’s care settings were also extracted, to identify incidents where age had not been recorded.

Data was recorded in a pro forma in Microsoft Excel. We asked MGPs to follow a protocol detailing data extraction, processing and anonymisation. They rechecked categorical information and removed identifiable details, before transferring the fully anonymised dataset to the research team.

Inclusion and exclusion criteria

RLC collated incidents involving IV fluid prescribing errors in a spreadsheet, double-checking the full dataset to ensure all were included. NICE guidelines were used to define IV fluids as ‘therapy to prevent or correct problems with fluid and/or electrolyte status’.[23] An attenuated dataset is available as S1 Dataset; an example incident with paraphrased narrative content is available as S1 Example incident.

To develop a full and authentic picture of IV fluid prescribing errors in practice, we included all incidents occurring across the five Trusts, including those where separate guidelines apply, such as neonatal care and diabetic ketoacidosis (DKA).

Two incidents were excluded: one involved a patient with hyponatraemia, which is a reporting trigger, but no error in IV fluid management was noted; the second involved heparinised saline used only to maintain central line patency, which fell outside the previously stated definition of IV fluids.

Study design and analysis

Our mixed methods approach involved:

  1. Reporting of incident characteristics with descriptive statistics
  2. Identification, classification and quantification of error types from narrative descriptions
  3. Thematic analysis of contributing factors, guided by Reason’s model of human error[24]

RLC initially quantified level of harm, age of patient affected, location of incident, and job role of reporter. Two authors (RLC and SMcV) then independently reviewed each incident, identifying types of error and developing a classification. A single incident could involve more than one error. Errors were defined as any reported deviation from accepted best practice at the time of study, with potential to cause harm. During the study period, practice in patients aged 0–16 was dictated by regional guidance, including a standardised chart,[25] and the 2007 NPSA safety brief.[26] Although NICE guidelines were not in place during the study period, and standards differed from recent recommendations, we felt there was value in contrasting types of error with current best practice. We therefore mapped error types to corresponding NICE recommendations.

The next stage was qualitative thematic analysis, employing Reason’s model of human error.[24] This ‘Swiss cheese model’ is commonly used for analysing prescribing errors,[27,28] working on the basis that harm occurs when deficiencies within a system align (Table 1).

thumbnail
Table 1. Incident analysis framework based on Reason’s model[24] (from Dornan et al[27], modified from Coombes et al[28]).

https://doi.org/10.1371/journal.pone.0186210.t001

Inter-rater reliability was addressed by two members of the research team (RLC and SMcV) independently coding data, using Reason’s four conditions as overarching categories. They resolved differences at all stages of analysis through discussion. By compiling and reviewing coded data relating to each aspect of Reason’s model, they jointly identified contributing factors underlying errors.

Ethics

The Proportionate Review Sub-committee of the East Midlands—Nottingham 2 Research Ethics Committee—approved the research (reference 15/EM/0353). Governance approval was granted by each of the five Trusts.

Results

Characteristics

From a dataset of 517 prescribing incidents, 40 reports relating to IV fluids were included in the analysis. IV fluid prescribing incidents were third most commonly reported, after those involving antimicrobials and paracetamol. Characteristics are summarised in Table 2. Twelve incidents led to patient harm, of which all but one was insignificant or minor in severity. The incident graded as moderate involved hypoglycaemia, but did not mention lasting harm after initial treatment. No incidents involved severe harm. Half were deemed to have potential to cause moderate or severe harm had they not been intercepted. 38% of incidents occurred in children aged 12–16. Incidents occurred in a range of clinical areas, 32% of which were outside specific paediatric settings. 40% were reported by nursing staff.

Types of errors

There were 70 errors in the 40 incidents, the principal types of which were: incorrect rate (27); inappropriate choice (17); and incorrect completion of fluid prescription chart (16) (Table 3).

Analysis of contributing factors

Contributing factors are summarised in Fig 1.

thumbnail
Fig 1. Factors contributing to IV fluid errors.

Examples are paraphrased; bold added by authors for emphasis.

https://doi.org/10.1371/journal.pone.0186210.g001

Latent conditions.

Organisational factors—non-specialist care settings (that is, those in which staff are not specifically trained in the care of children as, for example, paediatricians or paediatric nurses are), and doctors in a range of specialties prescribing for children—contributed to errors. Several involved failure to use paediatric charts and exceeding advised maximum rates of fluids, particularly in children looked after on adult wards. Errors arose when approaches used in adult practice were applied to children, exemplified by an adolescent being prescribed a litre of 5% dextrose over four hours. Reporters frequently noted prescribing errors occurring outside normal working hours. In some of these cases, not accessing support from a senior, supervising doctor, or an appropriate sub-specialist, was a factor.

Error-producing conditions.

Features of the patient, task, environment, or team were involved in most errors. Chronic illnesses, such as diabetes, renal disease, and metabolic disease, made management more complex. In these cases, prescribers often had difficulty negotiating additional protocols, whether patient-specific, as in metabolic conditions, or disease-specific, as in DKA. In some cases, there was perceived conflict between protocols: for example, a reporter noted that following a DKA protocol meant exceeding the advised maximum rate of fluids. Patients with rapidly changing clinical conditions were error-prone, such as when glucose was not included in the initial IV fluid prescription for a child with vomiting, leading to hypoglycaemia. A ‘task’ feature identified was prescribers frequently omitting body weights or rate calculations from charts. Sometimes this was because the requisite paediatric charts were not available.

Poor communication within teams contributed to errors. One incident describes a prescription being amended to correct hypokalaemia, but not being administered as the change was not communicated. In another, nursing staff gave fluids based on a verbal instruction which was later found to be different from the written prescription. Occasionally, incidents resulted from disagreement between staff. For example, an anaesthetist persistently requested that fluids be administered at an excessive rate despite a nurse’s concerns that this was incorrect.

Active failures.

Active failures occurred across the range of types described by Reason. Knowledge-based mistakes were seen in miscalculation of IV fluid rates. Often these resulted from unfamiliarity with the Holliday-Segar formula,[29] or difficulty applying it. More common were rule-based mistakes, in which doctors understood principles of prescribing fluids for children, but did not apply these appropriately in specific contexts. For example, a child with a urea cycle disorder was treated with IV fluids which, while usually appropriate, contained too much sodium and insufficient glucose. There were also examples of ‘unintended actions’ causing errors–slips, like prescribing fluids on the wrong patient’s chart, and lapses, such as forgetting to monitor. Occasionally, violations occurred, such as choosing to use an adult fluid balance chart—often to save time—or knowingly exceeding the maximum recommended rate of fluids.

Defences.

‘Defences’ refers to systems or staff actions to detect errors, prevent them reaching patients, or mitigate their effects. Nine errors were intercepted before fluids were given; most others were recognised before any significant harm occurred. By their nature, reported incidents relate to errors which have been picked up, and not those that go undetected. It follows that incidents contain information about successful defences, as well as missed opportunities to detect errors earlier.

Five errors were detected during mandatory pre-administration checks; 24 errors were missed during checking and detected after fluids had been started. The checking process appeared unreliable and depended heavily on vigilance of individual staff. Most errors were noted only when patients moved between wards (16 instances); their care was taken over by new staff (four); or their clinical condition changed (four).

Sometimes staff did not use intended safeguards, such as when they used adult rather than paediatric charts or failed to complete calculation guides incorporated within charts. In another case, a doctor and nurse bypassed pre-administration checks and administered IV fluids prior to the prescription being written. Some incidents indicated vulnerabilities in systems: one involved a patient being inaccurately weighed without any procedure for double checking; in another, the existence of two seemingly discrepant protocols made it difficult to identify unsafe practice.

The ability of staff to act as a defence depended not just on their vigilance but their level of expertise; some errors in complex patients went undetected until sub-specialist teams (e.g. paediatric metabolic team) reviewed them. Prescribers sometimes delayed seeking help from senior doctors or involving sub-specialists.

Discussion

Summary and context

Like adults in previous studies,[15] children were endangered by incorrect rates, inappropriate choices, and incorrect charting of fluids. Many specific types of error occurred, most unrelated to hyponatraemia or use of hypotonic fluids. Most errors seen relate to aspects of practice deemed key priorities in recent NICE guidelines.[2]

Interactions between individual, social, organisational, and environmental factors contributed to errors; knowledge deficits tended to be contextual rather than factual. To practise safely, prescribers had to negotiate challenging workplaces, navigate protocols, communicate effectively, use multiple resources, and correctly apply knowledge and rules. These findings resonate with those from earlier authors who showed prescribing to be a complex and inherently contextual process.[27,28,30] As reported before,[16] factors leading to errors were more likely to affect prescribers who were more junior, and less familiar with treating children.

Clinical audit and Northern Ireland paediatric admissions data[31] suggest that tens of thousands of intravenous fluid prescriptions were written during the period when the 40 errors occurred. This suggests that intravenous fluids are generally safe, but underlines the importance of the topic, given how frequently they are prescribed.

Strengths and limitations

A strength of the study was that it used a large, pre-existing critical incident dataset to gain insights into authentic clinical practice and drive improvements. We maintained rigour throughout. We adopted a comprehensive strategy to capture data. Experienced medicines governance pharmacists vetted incidents, both after initial reporting, and following data extraction. This made it more likely that all incidents were captured, and that categorical components such as level of harm were accurate. Two authors independently coded the dataset at the data analysis stage. Discussion of findings encouraged reflexivity and careful consideration of data limitations during interpretation. We only reported themes about which we reached consensus, and have presented examples of narrative data in support of these. All members of the research team, chosen to represent a range of disciplines, agreed on the final analysis.

We recognise the limitations of using critical incidents. Firstly, these are subject to under-reporting.[32] This may partly explain why only 40 incidents were reported in a time period when thousands of prescriptions were written. Reporting is also selective and motivations to report differ between staff groups.[33,34] The fact that, in our study, more incidents were reported by nurses than doctors, may reflect this. Reporting biases could have affected the type of incidents seen, and the underlying contributing factors identified.

Secondly, incident reports can be incomplete and of variable quality. Some fields had data missing, such as who had reported the incident. Other potentially useful information, such as the grade of doctor responsible, was not routinely reported. Furthermore, the way incident reports were written made it difficult to get information about some aspects of Reason’s model; for example, whether a doctor made a knowledge-based or rule-based mistake. Similarly, staff did not usually comment on contextual factors that led to errors, such as distractions or clinical pressures.

These factors limit the conclusions that can be drawn, particularly, from our quantitative data. Whilst these cannot be considered representative, they are nevertheless informative. Existing evidence (as described previously) is limited. Most comes from clinical audits focussing on limited aspects of fluid prescribing, such as use of hypotonic fluids. Given this limited evidence base, educators and quality improvers can use our findings about the many different types of errors seen to improve practice, pending more representative data from prospective research.

Qualitative analysis differs in that insights are drawn from the words within narrative descriptions, not how frequently events occur. By their nature, critical incidents contain information pertinent to patient safety. We used the rich evidence within these accounts to elicit factors contributing to errors. We were guided by the strength of the evidence presented, and its importance to safe fluid prescribing. In this way, the validity of our conclusions does not depend on representativeness. Our findings are transferable to other settings and contribute to research priorities identified by NICE.

Recommendations and conclusion

Our research recommendation is for prospective studies to advance the epidemiology of errors. Table 4 summarises educational recommendations. Undergraduate paediatric placements should teach fluid prescribing. The induction of all doctors who treat children, not just paediatricians, should teach how to prescribe fluids and make best use of information resources and clinical guidelines. Given the contextual nature of errors, learners need to practise prescribing, under supervision, and in context, before prescribing ‘solo’. Specific training should address special clinical situations, such as DKA or neonatal care. Interprofessional education, finally, could promote safe, collaborative practice and help nurses intercept errors.

thumbnail
Table 4. Recommendations for IV fluid prescribing education.

https://doi.org/10.1371/journal.pone.0186210.t004

Our clinical recommendations (Table 5) are that prescribing could be made safer in pressured clinical services by not treating children in adult wards,[35] using paediatric fluid balance charts with built-in prescribing safeguards, and ensuring clinical pharmacists are available and involved.[36] It is hoped this will replicate the positive impact they have had on prescribing other drugs.[37,38] Newer solutions such as electronic prescribing could offer opportunities to improve safety, by making safeguards more difficult to bypass. Critical incidents can also draw attention to specific systems improvements; for example, errors resulting from incorrectly recording patient weight could be prevented by a double checking system. Our study demonstrates the potential benefit from large-scale analysis of critical incidents; we recommend that, as well as being used locally, IV fluid prescribing incidents be collated, studied and shared more broadly.

thumbnail
Table 5. Recommendations for systems changes to improve IV fluid prescribing.

https://doi.org/10.1371/journal.pone.0186210.t005

Given the complex nature of the problem, it is unlikely any single measure will be fully effective. Complex interventions, incorporating some or all of the above measures, are most likely to succeed in making fluid prescribing safer.

Supporting information

S1 Risk matrices. Risk matrices used within Trusts to assign incidents level of actual and potential severity.

https://doi.org/10.1371/journal.pone.0186210.s002

(PDF)

S1 Dataset. Attenuated version of study dataset.

Narrative content has been removed to protect confidentiality of patients and staff involved.

https://doi.org/10.1371/journal.pone.0186210.s003

(XLSX)

S1 Example incident. Example of a critical incident, including paraphrased narrative information.

https://doi.org/10.1371/journal.pone.0186210.s004

(PDF)

Acknowledgments

The authors wish to thank the Northern Ireland Medicines Governance Team for extracting the data for this study. Thanks to Jenny Johnston for advice on writing the manuscript.

References

  1. 1. Moritz ML, Ayus JC. Prevention of hospital-acquired hyponatremia: a case for using isotonic saline. Pediatrics 2003;111(2):227–232. pmid:12563043
  2. 2. National Institute for Health and Care Excellence (NICE) (2015). Intravenous fluid therapy in children and young people in hospital: NICE Guideline [NG29]. Available at: https://www.nice.org.uk/guidance/ng29/resources/intravenous-fluid-therapy-in-children-and-young-people-in-hospital-pdf-1837340295109 [accessed May 2017]
  3. 3. Armon K, Riordan A, Playfor S, Millman G, Khader A. Hyponatraemia and hypokalaemia during intravenous fluid administration. Arch Dis Child 2008;93(4):285–287. pmid:17213261
  4. 4. Baker J, Armon K. The use of hypotonic fluids in paediatric practice. Arch Dis Child 2012;227(Suppl 1):A60–A61.
  5. 5. Banerjee J, Bhojani S, Khan A. Intravenous fluids and hyponatraemia—a hospital based retrospective cross-sectional study: comparing with the National Patient Safety Agency guidelines. Arch Dis Child 2010;95(Suppl 1):A52. http://dx.doi.org/10.1136/adc.2010.186338.115.
  6. 6. Junaid E. To National Patient Safety Agency or not to National Patient Safety Agency: an audit on the current trends in paediatric intravenous fluid prescribing for surgical patients. Arch Dis Child 2012;97:e9–e10.
  7. 7. Caldwell N, Williams L, Rackham O, Morecroft C. Do we still “treat” children with hypotonic intravenous fluids? Arch Dis Child 2015;100:e1.
  8. 8. Choong K, Kho ME, Menon K, Bohn D. Hypotonic versus isotonic saline in hospitalised children: a systematic review. Arch Dis Child 2006;91(10):828–835. pmid:16754657
  9. 9. Drysdale SB, Coulson T, Cronin N, Manjaly Z-R, Plysaena C, North A, et al. The impact of the National Patient Safety Agency intravenous fluid alert on iatrogenic hyponatraemia in children. Eur J Pediatr 2010;169(7):813–817. pmid:20012318
  10. 10. Freeman MA, Ayus JC, Moritz ML. Maintenance intravenous fluid prescribing practices among paediatric residents. Acta Paediatr 2012;101(10):465–468. pmid:22765308
  11. 11. McAloon J, Kottyal R. A study of current fluid prescribing practice and measures to prevent hyponatraemia in Northern Ireland’s paediatric departments. Ulster Med J 2005;74(2):93–97. pmid:16235760
  12. 12. Shukla S, Basu S, Moritz ML. Use of hypotonic maintenance intravenous fluids and hospital- acquired hyponatremia remain common in children admitted to a general pediatric ward. Front Pediatr 2016;4:90. pmid:27610358
  13. 13. Snaith R, Peutrell J, Ellis D. An audit of intravenous fluid prescribing and plasma electrolyte monitoring; a comparison with guidelines from the National Patient Safety Agency. Paediatr Anaesth 2008;18(10):940–946. pmid:18647271
  14. 14. Somarathna SS. Audit on intravenous fluid in children at a teaching hospital in North East England. Sri Lanka J Child Heal 2012;41(3):129–131.
  15. 15. Gao X, Huang K-P, Wu H-Y, Sun P-P, Yan J-J, Chen J, et al. Inappropriate prescribing of intravenous fluid in adult inpatients-a literature review of current practice and research. J Clin Pharm Ther 2015;40(5):489–495. pmid:26096723
  16. 16. Keijzers G, McGrath M, Bell C. Survey of paediatric intravenous fluid prescription: Are we safe in what we know and what we do? EMA—Emerg Med Australas 2012;24(1):86–97. pmid:22313565
  17. 17. Howell C, Patel B. An audit on the prescribing of intravenous fluids in paediatric patients. Arch Dis Child 2014;99(8):e3.
  18. 18. Davies P, Hall T, Ali T, Lakhoo K. Intravenous postoperative fluid prescriptions for children: A survey of practice. BMC Surg 2008;8:10. pmid:18541019
  19. 19. Davis D O’Brien MAT, Freemantle N, Wolf FM, Mazmanian P, Taylor-Vaisey A. Impact of Formal Continuing Medical Education. JAMA 1999;282(9):867–874. pmid:10478694
  20. 20. Rees P, Edwards A, Powell C, Hibbert P, Williams W, Makeham M, et al. Patient Safety Incidents Involving Sick Children in Primary Care in England and Wales: A Mixed Methods Analysis. PLoS Med 2017;14(1):e1002217. pmid:28095408
  21. 21. Williams H, Edwards A, Hibbert P, Rees P, Prosser Evans H, Panesar S, et al. Harms from discharge to primary care: mixed methods analysis of incident reports. Br J Gen Pract 2015;65(641):e829–e837. pmid:26622036
  22. 22. Arieff A, Ayus J, Fraser C. Hyponatraemia and death or permanent brain damage in healthy children. BMJ 1992;304:1218–1222. pmid:1515791
  23. 23. National Institute for Health and Care Excellence (NICE) (2013). Intravenous fluid therapy in adults in hospital: NICE Guideline [CG174]. Available at: https://www.nice.org.uk/guidance/cg174/resources/intravenous-fluid-therapy-in-adults-in-hospital-pdf-35109752233669 [accessed May 2017]
  24. 24. Reason J. Human error: models and management. BMJ 2000;320:768–770. pmid:10720363
  25. 25. Department of Health, Social Services and Public Safety (2010). Parenteral fluid therapy for children and young persons [Guideline]. Available at: http://www.ihrdni.org/303-060.pdf [Accessed May 2017]
  26. 26. National Patient Safety Agency (2007). Reducing the risk of hyponatraemia when administering infusions to children [Patient Safety Alert]. Available at: http://www.nrls.npsa.nhs.uk/resources/?EntryId45=59809 [Accessed May 2017]
  27. 27. Dornan T, Ashcroft D, Lewis P, Miles J, Taylor D, Tully M (2009). An in depth investigation into causes of prescribing errors by foundation trainees in relation to their medical education—EQUIP study [Report]. Available at: http://www.gmc-uk.org/FINAL_Report_prevalence_and_causes_of_prescribing_errors.pdf_28935150.pdf [accessed May 2017]
  28. 28. Coombes I, Stowasser D. Why do interns make prescribing errors? A qualitive study. Med J Aust 2008;188(2):89–94. pmid:18205581
  29. 29. Holliday MA, Segar WE. The maintenance need for water in parenteral fluid therapy. Pediatrics 1957;19:823. pmid:13431307
  30. 30. McLellan L, Tully MP, Dornan T. How could undergraduate education prepare new graduates to be safer prescribers? Br J Clin Pharmacol 2012;74(4):605–613. pmid:22420765
  31. 31. Department of Health (2016). Hospital Statistics: Inpatient and Day Case Activity Statistics 2015/16 [Report]. Available at: https://www.health-ni.gov.uk/sites/default/files/publications/health/hs-inpatient-day-case-stats-15-16.pdf [Accessed May 2017]
  32. 32. Sari AB-A, Sheldon TA, Cracknell A, Turnbull A. Sensitivity of routine system for reporting patient safety incidents in an NHS hospital: retrospective patient case note review. BMJ 2007;334(7584):79–79. pmid:17175566
  33. 33. Lawton R, Parker D. Barriers to incident reporting in a healthcare system. Qual Saf Heal Care 2002;11:15–18. pmid:12078362
  34. 34. Evans SM, Berry JG, Smith BJ, Esterman A, Selim P, O'Shaughnessy J, et al. Attitudes and barriers to incident reporting: a collaborative hospital study. Qual Saf Health Care 2006;15(1):39–43. pmid:16456208
  35. 35. Department of Health (2004). Getting the right start: National Service Framework for Children [Report]. Available at: http://www.nhs.uk/nhsengland/aboutnhsservices/documents/nsf%20children%20in%20hospitlaldh_4067251[1].pdf [accessed May 2017]
  36. 36. Staples A, Dade J, Acomb C. Intravenous fluid therapy—what pharmacists need to monitor. Hosp Pharm 2008;15:277.
  37. 37. Wang JK, Herzog NS, Kaushal R, Park C, Mochizuki C, Weingarten SR. Prevention of Pediatric Medication Errors by Hospital Pharmacists and the Potential Benefit of Computerized Physician Order Entry. Pediatrics 2007;119(1):e77–e85 pmid:17200262
  38. 38. Klopotowska JE, Kuiper R, van Kan HJ, de Pont A-C, Dijkgraaf MG, Lie-A-Huen L, et al. On-ward participation of a hospital pharmacist in a Dutch intensive care unit reduces prescribing errors and related patient harm: an intervention study Crit Care 2010;14(5):R174. pmid:20920322