Oxford Handbook of Clinical Surgery (4 edn)
Annotate
Cite
Duties of a doctor
The General Medical Council (GMC) lists the duties of a doctor in its document Good medical practice.1 The duties can be thought of under three headings (the 3 Cs): competency, communication, correctness (or probity).
Competency
Keep your professional knowledge and skills up to date.
Recognize the limits of your professional competence.
Perform an adequate assessment of the patient's conditions, based on the history and symptoms and, if necessary, an examination.
Arrange investigations or treatment where necessary.
Take suitable and prompt action when necessary.
Refer the patient to another practitioner when indicated.
Be willing to consult colleagues.
Keep clear, accurate, legible, and contemporaneous patient records that report relevant clinical findings, decisions made, information given to patients, and any drugs or other treatment prescribed.
Keep colleagues well informed when sharing the care of patients.
Provide the necessary care to alleviate pain and distress whether or not curative treatment is possible.
Prescribe drugs or treatment, including repeat prescriptions, only where you have adequate knowledge of the patient's health and medical needs. You must neither give or recommend to patients any investigation or treatment that you know is not in their best interests, nor withhold appropriate treatments or referral.
Report adverse drug reactions as required under the relevant reporting scheme and cooperate with requests for information from organizations monitoring the public health.
Take part in regular and systematic medical and clinical audit, recording data honestly, and respond to the results of audit to improve your practice, e.g. by undertaking further training.
Communication
Treat every patient politely and considerately.
Respect patients’ dignity and privacy.
Listen to patients and respect their views.
Give patients information in a way they can understand.
Correctness (or probity)
Make the care of your patient your first concern.
Respect the rights of patients to be involved in decisions.
Be honest and trustworthy.
Respect and protect confidential information.
Make sure your personal beliefs do not prejudice your patients’ care.
Act quickly to protect patients from risk if you have good reason to believe that you or a colleague may not be fit to practise.
Avoid abusing your position as a doctor.
Work with colleagues in the ways that best serve patients’ interests.
In an emergency, wherever it may arise, you must offer anyone at risk the assistance you could reasonably be expected to provide.
Confidentiality
Patients have a right to expect that information about them will be held in confidence by their doctors. Confidentiality is central to trust between doctors and patients. Without assurances about confidentiality, patients may be reluctant to give doctors the information they need in order to provide good care. The GMC states that if you are asked to provide information about patients, you must:
Inform patients about the disclosure or check that they have already received information about it.
Anonymize data where unidentifiable data will serve the purpose (this includes your surgical logbook).
Keep disclosures to the minimum necessary.
Keep up to date with and observe the requirements of statute and common law, including data protection legislation.
Daily practice
When you are responsible for personal information about patients, you must make sure that it is effectively protected against improper disclosure at all times (e.g. password-protected electronic files).
Many improper disclosures are unintentional. You should not discuss patients where you can be overheard or leave patients’ records, either on paper or on screen, where they can be seen by other patients, unauthorized health care staff, or the public. You should take all reasonable steps to ensure your consultations with patients are private.
Patients have a right to information about the health care services available to them presented in a way that is easy to follow and use.
Special circumstances
If in any doubt, contact your medical defence union for advice.
You must disclose information to satisfy a specific statutory requirement, such as notification of a known or suspected communicable disease. Inform patients about such disclosures, wherever that is practicable, but their consent is not required.
You must also disclose information if ordered to do so by a judge or presiding officer of a court. You should object if attempts are made to compel you to disclose what appear to you to be irrelevant matters.
You must not disclose personal information to a third party, such as a solicitor, police officer, or officer of a court, without the patient's express consent, except when:
The patient is not competent to give consent.
Reasonable efforts to trace patients are unlikely to be successful.
The patient has been or may be violent, or obtaining consent would undermine the purpose of the disclosure (e.g. disclosures in relation to crime).
Action must be taken quickly (e.g. in the detection or control of outbreaks of some communicable diseases) and there is insufficient time to contact patients.
References
Communication skills
Communicating with patients and relatives
When
During admission and before discharge.
On ward rounds.
During clinical examinations and procedures.
When the results of treatments are known and management changes.
In outpatient clinics.
Where
Maintain the patient's privacy. This is particularly important on an open ward. Knock on doors and close them after you. Draw the curtains round the bed. Ask a nurse to accompany you, particularly if you are explaining something complex or breaking bad news. They will have to answer the patients’ and relatives’ questions when you have left the ward or clinic room.
How
Know your facts. Are you giving the right diagnosis to the right patient? Are you equipped to consent a patient for the surgical procedure?
Sit at the same level as the person to whom you are talking, maintain appropriate eye contact, and introduce yourself.
Find out what the patient knows and what they are expecting.
Listen. The patient's own knowledge, state of mind, and ability to grasp concepts will dictate both how and how much you explain.
Tell the truth. Know your facts, be sensitive to what the patient may not want to know at this stage, and do not lie.
Avoid jargon. ‘Chronic’ may simply mean ‘longstanding’ to you; to most patients, it means ‘severe’.
Avoid vague terms. Try to describe risk quantitatively, ‘a 1 in a hundred chance’, rather than qualitatively, ‘a small risk’.
Check that the patient understands. Don't assume that they do.
Help the patient to remember. Use information booklets, draw diagrams, write instructions down.
Maintain a professional relationship. Never allow your personal likes, dislikes, and prejudices to hamper your clinical skills.
Breaking bad news
Is there a relative or friend whom the patient might wish to have with them, who may be a source of emotional support as well as being better able to retain information?
Know what options, if any, are available. If a cancer is inoperable, is chemotherapy planned? If an operation is cancelled, when is the next date?
Do not be afraid to stop to allow the patient time to gather their thoughts and emotions, and recommence at a later time.
Do not mistake numbness for calm acceptance and try not to take anger personally unless the bad news is actually your fault.
Communicating with nurses
Introduce yourself on arrival to the staff nurse in charge.
Establish early on which nurses are experienced. The help you get from them will be different from the questions you get from others.
In theatre, scrub nurses are not the enemy. Your inexperience is.
Try to remember all their names as they will remember yours.
Do ward work efficiently. Recognize how important it is for the smooth running of the ward that your ward rounds, note-keeping, prescriptions, and discharge letters are timely and accurate.
Let the nurses know when you are going for lunch, teaching, or sleep. If they can discuss problems now, it will save you being paged later.
Do an evening ward round to check on problem patients and drug requirements—your sleep is less likely to be constantly interrupted.
Communication with hospital doctors
Don't refer without first asking your consultant or registrar.
When making requests for clinical consultations, write a concise, but clear letter in the notes to the appropriate clinician.
When asked to see a patient, go the same day, write your opinion in the case notes, stating clearly what you recommend, and always discuss it with the seniors on your own firm.
If a preoperative patient is complex or has significant comorbidity, contact the appropriate anaesthetist. They will help you ensure that the patient is adequately prepared for surgery.
Communication with general practitioners (GPs)
The GP has usually looked after your patient for years and, however inspired your diagnostic or operating skills, they will be there to sort out all the complications that are hidden from you once the patient is discharged. They often know your consultant well. So think!
Telephone the GP in the case of a death of a patient, if you unexpectedly admit a patient, or to help with a difficult discharge.
Write useful, legible discharge summaries. What would you want to know if you were going to have to wait 4 weeks for the typed discharge letter to arrive—at an absolute minimum, the date and name of the operation, post-operative complications, and plan.
Keep clinic letters clear and concise.
Radiology and laboratory colleagues
Know exactly how the investigation will change your management.
If there is doubt about the correct investigation, telephone for advice.
Complete request forms correctly and include clinical data. It can make a big difference, particularly if you have requested the wrong test.
Administration
Introduce yourself to your consultant's secretary early, find out how they like things run, and then run things their way: they will usually have more than typing input on your reference.
Produce GMC, defence union, occupational health, holiday, and study leave paperwork with good grace. They are mostly legal requirements and being rude won't change that.
Evidence-based surgery
Summarizing simple data
This pattern of results is called a normal or Gaussian distribution: the curve is a symmetrical bell-shaped curve. Height, weight, age, serum sodium, and blood pressure (BP) are other examples of normally distributed data (see Table 1.1).
The mean is the same as the average: add up every result and divide by the number of results. The average Hb here is 11.1g/dL.
The standard deviation (SD) is a measure of how spread out the values are: result – mean = its deviation.
√((sum of deviations2/(sample size – 1)) = SD. Here SD = 1.6g/dL.
With normally distributed data, the mean ± 1 SD includes 68% of observations; ± 2 SD includes 95%; ± 3 SD includes 99%.
| Hb (g/dL) | No. of patients | Hb (g/dL) | No. of patients |
|---|---|---|---|
7–7.9 | 1 | 11–11.9 | 36 |
8–8.9 | 3 | 12–12.9 | 9 |
9–9.9 | 9 | 13–13.9 | 4 |
10–10.9 | 37 | 14–14.9 | 2 |
| Hb (g/dL) | No. of patients | Hb (g/dL) | No. of patients |
|---|---|---|---|
7–7.9 | 1 | 11–11.9 | 36 |
8–8.9 | 3 | 12–12.9 | 9 |
9–9.9 | 9 | 13–13.9 | 4 |
10–10.9 | 37 | 14–14.9 | 2 |
This pattern of results is called a skewed distribution. Post-operative blood loss (see Table 1.2), length of stay, and survival all show skewed distributions.
Don't use mean and SD to summarize skewed data.
The mean blood requirement, which is skewed to 8U of blood because of one outlier (*), is useful for planning budgets.
The best summary statistic for skewed data is the median (2U of blood) which is the value exactly halfway through the sample.
The interquartile range is what the middle 50% of observations were (1–2U here) and should be used instead of SD when summarizing skewed data.
Tests (see Table 1.3)
| Disease present | No disease | |
|---|---|---|
Test is positive | a | b |
Test is negative | c | d |
| Disease present | No disease | |
|---|---|---|
Test is positive | a | b |
Test is negative | c | d |
Sensitivity (a/(a+c))
A measure of how good the test is at correctly identifying a positive result (>98% is very sensitive). If a very sensitive test is negative, it rules the condition out (sign out).
Specificity (d/(b+d))
A measure of how good the test is at correctly identifying a negative result (>98% is very sensitive). If a very specific test is negative, it rules the condition in (spin).
Likelihood ratio
This is the chance that a person testing positive has the disease, divided by the chance that a person testing positive doesn't have the disease, or sensitivity/(1 – specificity). A likelihood ratio >10 is large and represents an almost conclusive increase in the likelihood of disease, <0.1 is an almost conclusive decrease, and 1 signifies no change.
Treatments and hazards (see Table 1.4)
| Outcome event | No outcome event | |
|---|---|---|
Exposure | a | b |
No exposure (control) | c | d |
| Outcome event | No outcome event | |
|---|---|---|
Exposure | a | b |
No exposure (control) | c | d |
Absolute risk reduction (ARR) a/(a+b) – c/(c+d)
This is the difference in the event rate between the control and the exposed group. It reflects the prevalence of a disease and the potency of a treatment or hazard.
Relative risk or risk ratio (RR) a/(a+b)/c/(c+d)
The event rate in the exposed group divided by the event rate in the control group. Used in randomized controlled trials (RCTs) and cohort studies. It is not affected by the prevalence of a disease.
Relative risk reduction (RRR) (a/(a + b) – c/(c + d))/(c/c + d)
ARR divided by the control event rate. It reflects disease prevalence.
Number needed to treat (NNT) 1/ARR
The number of people who must be treated to prevent one event.
Odds ratio
This is the odds of an exposed person having the condition divided by the odds of the control group having the condition. If the event is rare, it approximates to relative risk. Odds ratios are less intuitive than relative risk, but they are used because they are:
Usually larger.
Mathematically versatile.
Always used in case control studies and appear in meta-analyses of case control studies.
The basis of logistic regression analysis.
Statistical significance
Studies are designed to disprove the null hypothesis that findings are due to chance.
The p-value is the probability of a study rejecting the null hypothesis if it were true (a type I error), i.e. finding a difference where none exists.
Statistical significance is commonly taken as a less than 1 in 20 chance of this happening, i.e. p <0.05.
Power is the probability of detecting an association if one exists. Underpowered trials contain too few patients and may make type II errors, accepting the null hypothesis when it is false, i.e. finding no difference where one does exist.
95% confidence intervals, derived from the mean and SD, are the range of results predicted if the study were repeated 95 times.
Essentially incomplete data, usually due to variable lengths of follow-up. Common in surgical studies because 1) some patients will have been lost to follow-up and 2) patients will have shorter follow-up where they had operations more recently in a study.
Two methods used to calculate the percentage of study patients that survive a specified time after an operation when a study provides censored data.
Usually not curves. A linear graph, with percentage survival (or freedom from a complication) on the x-axis and time on the y-axis, which drops as each study patient dies (or gets the complication). If there are thousands of patients in the study, the curve is smooth. If there are very few, it is possible to see individual deaths/events as steps in the graph. Ideally, these graphs should have confidence intervals.
These reflect the precision of the study results. Narrow confidence intervals are better than wide ones because the confidence interval provides a range of values for the percentage survival (or odds ratio or other proportion) that has a specified probability (usually 95%) of containing the true value for the entire population from which the study patients were recruited. Always look for confidence intervals; they give you a ‘best case and worst case’ snapshot.
Essentially looking back from a group of patients with a known outcome (e.g. dead/alive) to see whether there were any predictors (e.g. age, recent myocardial infarction (MI)). Univariate analysis looks at single variables in turn. Multivariate analysis looks at a group of variables together; it is used to identify independent risk factors for an outcome. For example, age may be found to be a risk factor for post-operative death in univariate analysis, but that is because elderly patients are more likely to have other risk factors for post-operative death (e.g. recent MI). If age is not found to be an independent risk factor in multivariate analysis, it suggests that elderly people without other risk factors (e.g. recent MI) are not at higher risk of post-operative death.
Critical appraisal
Types of study
Studies appraising treatments can take several forms.
Randomized controlled trial (RCT)
Prospective study in which participants are allocated to control or treatment groups on a random basis. Gold standard for assessing treatment efficacy, but time-consuming and expensive to run.
Cohort study
Partly prospective study in which two cohorts of patients are identified, one of which was exposed to the treatment and one is the control group. They are followed over time to see the outcome. Cheaper and quicker than RCT and suitable for looking at prognosis, but prone to bias or false associations.
Case control study
Retrospective study in which patients with the outcome of interest are identified and paired with patients without the outcome of interest, and the exposure rates are compared. Cheapest and quickest way of looking for causation. Bias arises when patients are mis-classified as cases or controls.
Case series
A collection of anecdotes or case reports.
Systematic review
Differs from the traditional literature review by applying explicit, systematic, and reproducible methods to retrieve and appraise literature to answer a clearly formulated question. Large amounts of data are summarized and conclusions are more accurate.
Meta-analysis
A mathematical synthesis of the results of two or more primary studies, increasing the statistical significance of positive overall results. However, it reduces the ability of studies to demonstrate local effects.
Levels of evidence
Studies of treatment/hazard can be arranged in order of decreasing statistical validity.
Level 1a. Systematic review of RCTs.
Level 1b. High quality RCT with narrow confidence intervals.
Level 1c. All-or-none case series (either all patients died before treatment became available, but some now survive or some used to die, but now with treatment, all survive).
Level 2a. Systematic review with homogeneity of cohort studies.
Level 2b. Cohort study or low quality RCT.
Level 2c. ‘Outcomes’ research.
Level 3a. Systematic review with homogeneity of case control studies.
Level 3b. Individual case control study.
Level 4. Case series and poor quality cohort and case control studies.
Level 5. Expert opinion without explicit critical appraisal or based on physiology, bench research, or first principles.
How to appraise a paper
Answer these questions systematically. This information should all be stated explicitly within the manuscript.
How relevant is the paper?
Does the paper address a clearly focused, important, and answerable clinical question that is relevant to my patients?
How valid are the findings?
Was the paper published in an independent peer-reviewed journal?
Does the paper define the condition to be treated, the patients to be included, the interventions to be compared, and the outcomes to be examined?
Was a power calculation performed and is the power adequate?
Were all clinically relevant outcomes reported?
Was follow-up adequate?
Were all patients accounted for at the end of the study?
Was the appropriate study type selected and was the design appropriate?
Were the statistical methods described and were they appropriate?
Were the sources of error discussed?
Systematic reviews
Is the clinical question clearly defined and an acceptable basis for including or excluding papers?
Was the literature search thorough and were other potentially important sources explored?
Were trials appropriately included and excluded?
Was the methodological quality assessed and trials appropriately weighted?
RCTs
Were patients properly randomized?
Were patients treated equally apart from the intervention being studied?
Was analysis on an intention-to-treat basis?
Are confidence intervals narrow and not overlapping?
Case control studies
Were patients correctly classified as case or control?
Were all patients accounted for at the end of the study?
How important are the results?
Were the results statistically significant?
Were the results expressed in terms of numbers needed to treat and are they clinically important?
How applicable are the findings?
Were the study patients similar to mine?
Is the treatment feasible within my practice: is information on safety, tolerability, efficacy, and price presented?
Audit
What is audit?
National Institute for Health and Clinical Excellence (NICE) defines clinical audit as ‘a quality improvement process that seeks to improve patient care and outcomes through systematic review of care against explicit criteria and the implementation of change. Aspects of the structure, processes, and outcomes of care are selected and systematically evaluated against explicit criteria. Where indicated, changes are implemented . . . and further monitoring is used to confirm improvement in health care delivery.’
Why do it?
Clinical audit is currently seen as the most effective way of assessing routine health care delivery and the basis of improving outcomes.
All hospital doctors are required to fully participate in clinical audit (NHS Plan, Department of Health, 2000).1
The GMC advises that all doctors ‘must take part in regular and systematic medical and clinical audit . . . Where necessary, you must respond to the results of audit to improve your practice’.
How to do it
Audit of outcome or process can be divided into five stages: each stage needs to be carefully planned to produce a clinically effective audit.
Preparing for audit
Choose a topic and define the purpose of the audit. One option is to identify (by consulting patients and clinicians) a potential problem that may involve high costs or risks for which there is good evidence to inform standards and that may be amenable to change. NICE stresses the importance of identifying skills and resources to carry out the audit.
Selecting audit criteria
Audit can assess process or outcome.
Define the patients to be included.
Criteria to assess performance should be derived from the available evidence, e.g. trials, systematic reviews, society guidelines, or clinician consensus.
Benchmarking prevents unrealistically high or low targets.
Measuring performance
This is about collecting data. Identify patients or episodes from several sources (e.g. operating room logbooks and patient administration system (PAS)) to avoid missing patients because of incomplete data. Electronic information systems can improve data collection. Training dedicated audit personnel can improve the process further.
Making improvements
Identify local barriers to change, develop a practical implementation plan, which should involve several interventions (practice guidelines, education, and training). Clinical governance programmes should provide the structure.
Sustaining improvements
Repeating the audit to assess improvements is also called closing the audit loop. Alternatives such as critical incident review may be effective.
Measuring surgical performance
Rationale
The Kennedy report on the enquiry into perioperative deaths in paediatric cardiac surgery at Bristol Royal Infirmary stated that ‘Patients must be able to obtain information as to the relative performance of the Trust . . . and consultant units within the Trust’. The idea that every patient has the right to expect their surgery to be performed by a surgeon whose results are not statistically worse than average is widely held.
League tables ranking cardiothoracic surgeons on the basis of surgeon-specific mortality data have been published by the government.
Surgeons can be ranked using a number of other outcomes.
Ranking should incorporate a system that accounts for differences in case mix, i.e. risk stratification, so that surgeons who operate on sicker or more complicated patients are not unfairly penalized.
Risk scoring systems
Examples include:
Euroscore and Parsonnet scores for predicting operative mortality in cardiac surgical patients.
Apache II scores for intensive care patients.
POSSUM (Physiologic and Operative Severity Score for the enUmeration of Mortality and morbidity)—variants exist for vascular and colorectal surgery.
Presenting results
The aim of presenting performance data is to distinguish between normal variation between surgeons or institutions and significant divergence. There are three main ways of doing this:
Average outcome over a given time frame.
Ranking or league tables of surgical mortality or other complications; the data may be crude or risk-stratified (i.e. taking into account the case mix).
Survival plots which may also be crude or risk-stratified.
Standardized mortality ratio plots.
Volume and outcome control charts.
Funnel plots (see Fig. 1.1).
Spectrum plots.
Performance trends over time.
Cumulative summation charts (CUSUM).
Variable life-adjusted display charts (VLAD), risk-adjusted CUSUM.
Funnel plot of mortality data for 50 cardiac surgeons. The arrow marks an outlier with mortality outside 95% confidence intervals (CI).
References
Consent
Legal aspects
Successful surgery depends on a relationship of trust between the patient and doctor. The patient's right to autonomy must be respected, even if their decision results in harm or death. This right is protected by law.
A doctor performing a procedure on a patient without their consent can be found guilty of battery.
A doctor who has failed to give the patient adequate information to allow them to give informed consent can be found guilty of negligence.
No adult in the United Kingdom (UK) can legally consent to surgery on behalf of another adult. It is important to involve relatives, particularly where patients are unable to consent, but their wishes are not legally binding and do not form part of the legal consent.
Obtaining consent
The key to good consenting is good communication (see 
p. 4). It may be necessary to use a translator and some Trusts will not accept consent gained by using patients’ relatives as translators. GMC guidelines state ‘If you are the doctor providing treatment or undertaking an investigation, it is your responsibility to discuss it with the patient and obtain consent’. In practice, this may be done verbally by the consultant or registrar in clinic, or on a ward round with a house officer obtaining written confirmation later. Consent must be given freely: patients may not be put under duress by clinicians, employers, police, or others to undergo tests or treatment. Declare any potential conflicts of interest. The amount of information should be sufficient to allow a mentally competent patient to make an informed decision. It will vary according to the individual, the nature of the condition, the complexity of treatment, and risks involved. It is unacceptable to limit the amount of information on the basis that it may cause distress, but be sympathetic to the patients’ needs. Consent must be obtained for taking photographs for teaching or publication, and taking samples for research.
Informed consent
There are five aspects that the patient must understand to give informed consent:
The reason for carrying out the procedure. The patient needs to understand the nature of their illness and its prognosis.
What the procedure involves. Where and how long is the scar; what is being removed; what prosthesis will be implanted; will there be drains?
The risks of the procedure. Specific to the procedure (e.g. stoma, limb dysfunction) and in general (e.g. anaesthesia, bed rest, deep vein thrombosis (DVT)).
The benefits of the procedure. Improvement in symptoms or prognosis or purely diagnostic.
Alternatives. Including conservative treatment, with their advantages and disadvantages.
Modes of consent
Implied consent. The patient is presumed to consent to minor procedures, e.g. X-rays, phlebotomy, by cooperating with ward procedures.
Express written consent. Whenever possible, this should be obtained for all patients undergoing procedures involving an anaesthetic, complex treatments with significant risks and side effects, or as part of research. Written consent is not legal proof that adequate consent was obtained at the time the document was signed.
Express verbal consent. Should be obtained when it is not possible to get written consent, witnessed by an independent health care professional, and documented in the notes accordingly, or for simple procedures with minimal risk of harm.
Special considerations
Emergencies
When consent cannot be obtained, you may provide emergency medical treatment, provided it is limited to what is needed to preserve life. However, you must respect any valid advance refusals that you know about or that are drawn to your attention.
Mentally incapable patients
No adult in the UK can legally consent to surgery on behalf of another adult. Assess the patient's competence to make an informed decision. If unable to decide, and provided they comply, treatment may be instigated that is judged to be in their best interests. Otherwise, treatment may be carried out under the Mental Health Act 1989. Controversial and non-therapeutic treatments (e.g. sterilizations) require court approval.
Advance statements/living wills
Advance statements made by patients before losing the capacity of informed consent must be respected, provided the decision is applicable to the present circumstances and there is no reason to believe that they may have changed their minds. The known wishes of the patient should be taken into consideration if an advance statement is unavailable.
Children
Over 16s are regarded as young adults and have capacity to decide.
Under 16s may give their own consent if they are judged to understand what is involved.
Unlike adults, where a competent child refuses treatment, a person with parental responsibility (except in Scotland) or a court may authorize treatment if deemed in the child's best interests.
If the parents refuse treatment deemed in the child's best interests, you are not bound by this and may seek a ruling from the court.
Emergency treatment may be instigated without consent in a similar manner to that in adults.
Pregnancy
The right to autonomy applies equally to pregnant women. It includes the right to refuse treatment that is intended to benefit the unborn child.
Death
Confirming death
There is no legal definition of death in the UK. It is generally regarded as the cessation of circulation and respiration.
Clinically, there is:
No respiratory effort, denoted by the absence of breath sounds on auscultation over 1min.
Absence of a palpable pulse and heart sounds over 1min.
No response to painful stimuli, e.g. sternal or supraorbital rub.
Fixed dilated pupils (beware drugs such as atropine).
If there is doubt, perform an electrocardiogram (ECG).
Hypothermia (core temperature <34°C) must have been corrected.
Brain death/brainstem death
The concept of brain death has arisen from the advances in intensive therapy, and the ability to maintain cardiac and respiratory function artificially in patients who have sustained severe irreversible brain damage. Brain death is defined as the ‘irreversible cessation of all functions of the entire brain, including the brainstem’.1 This, alongside the traditional definition, is taken to equate to death in the UK, United States of America (USA), Australia, and many other countries. In order to diagnose brain death, a number of strict criteria must be met.
An identifiable cause for the brain death must be established, e.g. severe head injury/intracerebral bleed, and
Other causes, including central nervous system (CNS) depressants, hypothermia, metabolic and endocrine disturbance, need to be excluded, and
The patient is unable to breathe spontaneously despite adequate CO2 drive (i.e. PaCO2 >6.7kPa), and
The following brainstem reflex tests, performed by the consultant in charge (or deputy of 5y registration) and another suitably experienced doctor, have been failed on two separate occasions, usually 24h apart:
Both pupils are fixed and unresponsive to light (oculomotor nerve).
Corneal reflexes are absent (trigeminal nerve).
Vestibulo-ocular reflexes are absent (absent eye movements when 20mL of ice-cold water is injected into each ear with tympanic membranes visualized beforehand) (vestibulo-cochlear nerve).
Absent motor responses to painful stimuli in the distribution of the cranial nerves in the absence of neuromuscular blockade (spinal cord injury may ablate peripheral motor responses).
Absence of respiratory effort when disconnected from the ventilator despite a PaCO2 >6.7kPa (↑ in chronic obstructive pulmonary disease (COPD)).
Absent gag and cough reflex upon pharyngeal and endotracheal stimulation.
Coroners
It is always wise to discuss with the consultants involved if there is any reason to discuss cases with the coroner's officer. Poor quality information can lead to death certificates being returned or the coroner becoming involved unnecessarily.
In-hospital deaths must be discussed with the coroner's officer if:
Death has occurred during an operation.
Death occurred before recovery from anaesthetic.
More than 14 days have elapsed since the patient last saw a doctor.
There is doubt about the cause of death.
Death is thought to be suspicious (e.g. caused by overdoses of prescribed substances, medical error, suicide).
Certifying death
Documenting in the medical notes
If you are asked to ‘certify’ a patient, first confirm death (see p. 16):
Document the date and time that death was pronounced.
Document your examination.
Document the causes of death as they will appear on the death certificate if these have been decided. If in doubt, always speak to the consultant.
The death certificate
This can be issued by anyone with full medical qualifications who looked after the patient during their last illness, or where referral to the coroner has been made and permission to issue the certificate has been granted.
Write legibly. The record is retained by the relatives and illegible or incomplete certificates may be rejected by the funeral director.
Part I. The cause of death. Events leading to Ia are listed in Ib and Ic.
Part II. Conditions that contributed to, but did not directly cause death.
General terms like heart failure and sepsis may not be accepted.
Cremation forms
These forms vary slightly between regions, but certain rules always apply.
There are two parts. The first is filled in by a doctor who attended the patient during the illness leading up to death, the second by an independent clinician who has been fully registered for at least 5y.
They should not be issued if the cause of death is not established.
It is the responsibility of the issuing doctor to ensure that they have seen and identified the person after death, and that there are no radioactive implants or pacemakers present.
Post-mortems
A coroner's post-mortem is required for suspicious deaths, but is most commonly performed where the Coroner's Office has ‘taken’ a case where the cause of death is uncertain, or may be related to surgery or interventions. The consent of relatives is not necessary to proceed.
A hospital post-mortem may be carried out with the consent of relatives to investigate other deaths. In 60% of post-mortems in one series, new diagnoses that would have substantially changed management were found—they are a vital part of audit.
References
End-of-life issues
Do not resuscitate (DNR) orders
A DNR order should be considered when the frailty, comorbidity (e.g. inoperable disseminated malignancy, multiple organ failure), maximal medical treatment, or advanced age of a patient means that any attempt at cardiopulmonary resuscitation (CPR) in the event of a cardiac or respiratory arrest will be futile. DNR decisions should be reached on a case by case basis: a blanket ‘do not resuscitate’ policy based on a specific patient group, such as elderly patients, is unacceptable. An 84-y-old patient who was an appropriate candidate for cardiac surgery is an appropriate candidate for CPR post-operatively, whereas a 72-y-old patient undergoing palliative care for end-stage hepatorenal failure is probably not.
Never make a DNR decision without discussing it with a consultant.
Patients and, where appropriate, their relatives must be involved.
Document the clinical reasons for the DNR order and state explicitly whether ‘full active medical management’ is to be continued: DNR orders do not always include withdrawing treatment. Discuss each case with the nurses involved.
Complete the appropriate documentation and review process, which varies from Trust to Trust, and make sure the nursing staff are fully aware so that they do not call the arrest team when the patient dies.
Euthanasia
Euthanasia is the painless termination of life at the request of the patient concerned. In the UK, it is illegal to administer any drug to accelerate death, irrespective of how compassionate the motive may be. Withdrawing futile treatment is not euthanasia. UK law states that the intention to kill is malicious and such action would be classified as murder. Terminally ill people and the parents of terminally ill or severely disabled children may have several reasons for requesting euthanasia. Effective palliative care, counselling, and multidisciplinary support should be able to address most of these reasons, which include:
Pain.
Disability.
Disfigurement.
Depression.
Fear of being a burden, being unable to cope.
Palliative care
Palliative care is surgical, medical, and nursing care aimed specifically at relieving the problems associated with terminal conditions when the possibility of cure has been abandoned. Palliative care is delivered by palliative medicine and nursing specialists and can take place in the community or in residential care settings. Refer early: palliative care beds are limited and acute surgical wards are rarely the best places for dying patients. Palliative care physicians specialize in:
Control of symptoms, including pain, anorexia, nausea and vomiting, confusion, dysphagia, dyspnoea, incontinence.
Psychological aspects of terminal illness.
Bereavement.
Suicide
The suicide rate in the UK is currently 12.5 per 100 000.
Patients at risk
The recently bereaved.
Cancer patients have a five times increased risk.
Men over 55y with oral cancer and a history of alcohol abuse.
Women of any age, often suffering from gynaecological or breast cancer. (In both of these latter groups, the treatment of the disease involves disfigurement and a change of body image.)
Action
Patients about to undergo disfiguring surgery for any reason should be counselled carefully in the period after confirmation of the diagnosis and before surgery. Doctors should discuss all treatment options and implications clearly. The support of a ‘mastectomy counsellor’ or ‘stoma therapist’ is invaluable. Post-operatively:
Look for symptoms of depression, including low mood, tearfulness, anorexia, early morning waking, suicidal thoughts, especially in long-term patients.
Do not discontinue antidepressant medication.
Ensure that arrangements for discharge include community nursing support and that the GP is aware of the patient's state of mind.
Organ donation
When brain death is established, organ donation should be considered for all patients who are under 75y of age with no history of malignant disease or major untreated sepsis.
All donors should be tested for human immunodeficiency virus (hiv), hepatitis B and C, herpes simplex virus (HSV), and cytomegalovirus (CMV).
Organ donation is usually coordinated by regional transplant teams.
The body should be identified and next of kin contacted.
If, despite reasonable attempts, the identity of the corpse or next of kin remains unknown, the body becomes the property of the health authority.
If a donor card is present, it is reasonable to assume that the deceased wished to donate his organs and the transplant team can proceed.
If relatives are identified and do not wish organ donation to proceed, even though there is a donor card, their wishes must be respected.
Relatives should be asked to act as agents in expressing what they believe to be the wishes of the patient. Ideally, the person seeking permission should be someone whom they already know. This may be the consultant in charge, but, on occasion, a senior staff nurse, chaplain (or other religious figure), or the family GP may be more appropriate.
In the case of accidental deaths, the coroner's permission should be sought before proceeding.
Clinical governance
Clinical governance is the system through which National Health Service (NHS) organizations are accountable for continuously improving the quality of their services. Clinical governance involves setting standards, performance monitoring, and reporting systems at national, institutional, and personal levels. Risk management is an integral part of clinical governance: it is the systematic identification and avoidance of risks associated with any procedure.
Setting standards
In addition to conventional clinical evidence and guidelines, the following organizations have a responsibility for setting standards in health care.
National Service Frameworks (NSFs)
NSFs are long-term national strategies for improving specific areas of care, produced by government after consultation with clinicians. They set standards and establish methods of delivering them. NSFs have been published for coronary heart disease and cancer.
National Institute for Health and Clinical Excellence (NICE)
NICE is the governmental organization responsible for setting standards by reviewing the best available evidence and publishing guidelines. Local authorities are obliged to fund interventions recommended by NICE, but NICE guidance does not overrule individual clinical decision-making. Currently, NICE produces three kinds of guidance:
Technology appraisals. Guidance on the use of new and existing medicines and treatments within the NHS in England and Wales.
Clinical guidelines. Guidance on the appropriate treatment and care of people with specific conditions within the NHS in England and Wales.
Guidance on whether interventional procedures used for diagnosis or treatment are safe enough and work well enough for routine use in England, Wales, and Scotland.
Postgraduate Medical Education and Training Board (PMETB) and professional organizations
The PMETB was set up in 2003 to develop a single, unifying framework for postgraduate medical education (PGME) and training across the UK.
Medical ethics, undergraduate and pre-registration medical education, and fitness to practise remain the responsibility of the GMC.
Accreditation and the approval of basic surgical training remain the responsibility of the Royal Colleges of Surgeons.
Higher surgical training is supervised by the Joint Committee on Higher Surgical Training (JCHST) and the specialist advisory committees (SACs).
Performance monitoring
Healthcare Commission (replaces CHI and the Audit Commission)
The Healthcare Commission is a new body that has been set up to help improve the quality of health care. It will do this by providing an independent assessment of the standards of services, whether they are provided by the NHS, independent health services, or voluntary organizations. It provides an independent second stage of complaints assessment, assesses the arrangements in place to promote public health, and acts as the coordinating inspectorate in relation to health care.
National Confidential Enquiry into Patient Outcomes and Death (NCEPOD)
NCEPOD (which used to be the National Confidential Enquiry into PeriOperative Deaths) is an organization independent of the Department of Health (DOH) and the professional associations, although it receives over 85% of its funding from the DOH via NICE. It stopped collecting data on all deaths within 30 days of surgery in 2002 when its remit was extended to cover all medical and surgical deaths. Data is now collected by local reporters in response to specific areas of research.
National Patient and User Survey
The Healthcare Commission has carried out five national surveys involving over 500 institutions and 300 000 patients. The results are disseminated to health care providers and governmental agencies to inform strategy.
Audit
See p. 12.
Revalidation
The purpose of revalidation will be to create public confidence that all licensed doctors are up to date and fit to practise. Currently, consultants and general practitioners undergo regular appraisal: this will be included within revalidation which is yet to be introduced.
Reporting systems
Critical incident reporting
Critical incident reporting was initially voluntary and anonymous. Incidents perceived to have exposed patients or staff to actual or potential risk were reported on forms that would be sent to responsible individuals in each directorate, serious adverse events being discussed at regular meetings with the clinical director. Now effective critical incident reporting systems are a requirement for CNST (criminal negligence scheme for Trusts) insurance and are assessed by the Healthcare Commission on Trust visits.
Complaints
Patients make formal complaints about treatment or clinicians to the hospital concerned: the new NHS complaints procedure, which was reformed in 2002 to address concerns that it was fragmented, complex, and insufficiently independent, still stresses that wherever possible, issues should be resolved locally. Effective use of the patient advice and liaison services (PALS) and independent complaints advocacy service (ICAS) should mean that only serious complaints are referred to the second level, which is handled by the Healthcare Commission which currently handles 3000–5000 complaints per annum.
Whistle-blowing
Trusts are required to have a whistle-blowing policy to enable individual staff members to express concerns about treatment and to protect them from reprisals. The policy must include a mechanism for investigating and acting on such claims.
