Professionalism, stats Flashcards

Question

Cultural competence

Answer 1

“an awareness of cultural diversity and the ability to function effectively, and respectfully, when working with and treating people of different cultural backgrounds. Cultural competence means a doctor has the attitudes, skills and knowledge needed to achieve this.” “Cultural competence is a set of congruent behaviours, attitudes, and policies that come together in a system, agency, or among professionals and enable that system, agency, or those professionals to work effectively in cross-cultural situations.”

Answer 2

‘The need for doctors to examine themselves and the potential impact of their own culture on clinical interactions and healthcare service delivery. The commitment by individual doctors to acknowledge and address any of their own biases, attitudes, assumptions, stereotypes, prejudices, structures and characteristics that may affect the quality of care provided. The awareness that cultural safety encompasses a critical consciousness where healthcare professionals and healthcare organisations engage in ongoing self-reflection and self-awareness and hold themselves accountable for providing culturally safe care, as defined by the patient and their communities.’

Answer 3

- respect and understanding - culturally tailored communication - patient centred practice - partnership issues specific to anaesthesia and pain medicine: - patient vulnerability - cultural understanding of pain - culturally important situations including childbirth, death and dying, blood products, organ/tissue transplantation

Answer 4

* Improved quality, safety and experience of care for individual patients. * Improved health and equity for the population. * Best value for the available resources.

Answer 5

* Safety * Timeliness * Efficiency * Efficacy * Equitability * Patient-centredness

Answer 6

“an organised process designed to ensure the maintenance of a desired level of safety and quality in a service or product”

Answer 7

“an iterative process to continuously improve the safety and quality of care provided to patients”

Answer 8

- planning - implementation: collection of data and its analysis, review of results, and action to be taken - review - setting/revising standards/policies

Answer 9

Aims: - correct medication to the correct patient, in the correct dose, by the correct route, at the correct time - accurate recording - minimise opportunities for misuse Anaesthetists should - have a good understanding of the drug's PK, PD, PC including complications - have a good understanding of the laws, systems and processes surrounding prescription and administration - have a good awareness of factors contributing to drug error and take steps to manage them - collaborate with pharmacists and safety groups - audit regularly

Answer 10

For patients treated in Australia, if no person is able to give consent, then treatment can proceed only if all of the following are satisfied: - It is in the patient’s best interests. - Reasonable steps have been taken to ascertain the views of the patient. - The doctor believes that it would have been the patient’s choice had they been competent to do so. - The doctor takes into account the views of other persons with a genuine interest in the welfare of the patient. - Any delay is likely to be detrimental to the patient.

Answer 11

Fatigue is the subjective feeling of the need to sleep, an increased physiological drive to fall asleep and a decreased state of alertness. - It has been demonstrated to impair vigilance and accuracy of response - Decreased performance of motor and cognitive functions in a fatigued anaesthetist may result in: -- impaired judgement -- late and inadequate responses to clinical changes -- poor communication/empathy -- increased errors

Answer 12

- planning: avoid undertaking clinical duties if physical or mental fatigue, stress or ill health, alone or in combination, might interfere with safe patient care - awareness: performance may be impaired, recognise and respond to symptoms in self and others - risk mitigation: naps, eating, caffeine, light exposure - recovery: naps, re establish normal activities - long term strategies: leisure activities, exercise, rest and sleep, recreation leave, work breaks

Answer 13

- surgery should not be started after 10pm unless it is a life-, limb or organ-threatening emergency - elective surgical list should have an expected completion time of no later than 10pm - maximum work 16 hours in a 24-hour period - roster should allow for adequate breaks - roster should allow for at least 8 hours before the commencement of the next shift - management plan for clinician non availaibility due to fatigue: expense of CST - staffing should allow regular rec leave - rosters should provide enough lead time to plan recreational activities - forward-rotating shifts (mornings - evenings - nights) are associated with the least disturbance to normal sleep patterns - should have enough shift overlap for safe handover - education, rest facilities, safe transport - monitor fatigue related incidents

Answer 14

Prolonged absence – any absence from clinical anaesthesia exceeding twelve months in duration. Some anaesthetists may require a return to anaesthesia practice program after shorter durations of absence

Answer 15

1. distress due to a variety of triggers without evidence or signs of substance misuse 2. substance misuse of non-anaesthesia prescription or recreational drugs including alcohol 3. out-of-work criminal charges related to illicit substance possession or misuse 4. more subtle or increasing signs of distress and mounting evidence of anaesthesia substance misuse 5. intoxicated or impaired colleague with direct evidence of using anaesthesia or nonanaesthesia drugs in the workplace - reports of potentially impaired anaesthetists should be supported by credible evidence and investigated in a timely manner with confidentiality - patient outcomes, such as higher report rates of inadequate pain relief or critical incidents - decline in clinical performance, including documentation errors or inconsistent care - behavioural changes, such as increasing isolation, erratic behaviour, frequently leaving the operating theatre, absenteeism, or mood swings - physical or cognitive signs of impairment, such as slurred speech or frequent fatigue

Answer 16

- stress, poorly functioning teams, inadequate support at work or home - pain, insomnia, or mental health disorders - perfectionist tendencies, maladaptive coping strategies, and competition for employment and career advancement - social and emotional trauma, experiences of racism, discrimination and harassment

Answer 17

- This is a crisis requiring immediate action - Immediately ensure any relevant patient is safe and that the practitioner is medically safe. - Arrangements should be made for cover for the practitioner’s patients/list. - An intervention interview needs to be urgently convened - Mandatory notification to the regulator is required. - Facilities should consider having a drug and alcohol policy which allows for drug screening on request preferably with an independent pathology facility A practitioner should then expect to provide a contemporaneous urine sample and/or hair sample for drug screening if requested, with attention to direct supervision and chain of custody requirements - It should be recognised that the practitioner will be under significant distress. Facilitated involvement of their partner or other suitable support person, GP, Addiction Services or other available support services can be lifesaving. - As the risk of self-harm in this period is particularly high the practitioner should not leave the facility alone and should be supported.

Answer 18

A measurable characteristic of a population (eg the average depth to the epidural space in all pregnant women in Australia). It has a fixed and typically unknown value

Answer 19

A measurable quantity of a sample (eg average depth to the epidural space in all pregnant women at Canberra Hospital in November 2025)

Answer 20

Mathematical quantity calculated from the measured sample , used to estimate the population parameter

Answer 21

Denotes data is symmetrical on either side of the midpoint. Has some mathematical connotations regarding dispersion. When graphed, is symmetrical + bell - shaped. Allows for more powerful statistics tests and assumptions mean = median = mode

Answer 22

Continuous and quantitative. → ejection fraction

Answer 23

Continuous and quantitative, but can only reasonably be expressed in integers → number of coronary artery grafts

Answer 24

Categorical data, where there is an implied rank order. → NYHA scores, Verbal pain.

Answer 25

Categorical, with no implied rank. → Adrenaline, Noradrenaline, Milrinone for CABG

Answer 26

most occurring response

Answer 27

average for normal data set

Answer 28

midpoint of data set from lowest to highest

Answer 29

highest - lowest

Answer 30

25th centile to 75th centile (ie middle 50%)

Answer 31

only for normal data "spread of the statistical data from the mean or average position" measured in same units as the mean = square root of variance Shows absolute variability of data Used when describing the variability of the actual sample data Pro: very accurate in normally distributed data Con: cannot directly compare data with means of different magnitudes/scales

Answer 32

SD/root of n Indicates precision of the sample mean, i.e. how far is the sample mean from the population mean. The 95% confidence interval for the population mean = sample mean ± 2 SEM Pro: can compare means of different populations, describes the true sample population Con: often misused when substituted for SD

Answer 33

comparing 2 samples and checking for difference OR estimating probability that sample reflects greater population

Answer 34

Patient Intervention Comparitor Outcome

Answer 35

Probability of observing the data obtained (or data more extreme from the prediction of the null hypothesis) if the null hypothesis is true P-value for statistical significance is defined as P < 0.05 If the null hypothesis is true and the sample means are not different, a difference between the sample means at least as large as that observed in the first study would be observed 5% of the time The conventional value for statistical significance (P < 0.05) should always be viewed in context A P-value close to this arbitrary cut-off point should perhaps lead to the conclusion that further work may be necessary before accepting or rejecting the null hypothesis

Answer 36

Confidence interval = range of sample data which includes an unknown population parameter, eg mean Ie the mean of the sample group is known, but not the mean of the population, therefore a range is given in which the population mean likely exists It does not indicate that 95% of the sample values occur in that range Helps to navigate the uncertainty of how well a sample estimates a value for an entire population; indicates likely magnitude of the effect being investigated and the reliability of the estimate Narrower CI is more precise If an investigation is repeated numerous times, the CI 95% generated will contain the population mean 95% of the time It does not mean there is a 95% chance that a specific interval contains the true value - that is a confidence level Increasing the confidence level (e.g., 95% to 99%) while holding the sample size and variability constant, will widen the confidence interval and vice versa CI helps to interpret the P-value - they should always be quoted together because a higher P-value will generate wider CI -> ie less precise If a CI includes the null value, the data do not support a meaningful difference between the population means Null value = value obtained when there is no difference between groups - eg difference between means = 0 Significance: If p < 0.05, the confidence interval will not contain the null hypothesis value (e.g., 0 for mean difference) Non-Significance: If p ≥ 0.05, the confidence interval will contain the null hypothesis value. When the lower and upper limits are both positive or both negative - depending on direction - then the difference is significant

Answer 37

Occurs when the null hypothesis is true, but a p-value less than 0.05 (or some other pre - specified level) is obtained. ie False positive, the result occurred purely by chance The risk of this can never be completely eliminated

Answer 38

Occurs when the null hypothesis is actually false (ie there is actually a difference between 2 groups) but statistical significance has not been achieved. ie False negative Main cause is inadequate sample size Can also be seen when an expected event (eg mortality) occurs less frequently then planned Small samples, non parametric tests

Answer 39

Calculating Power Power = 1 – β x 100% * Power calculations are used to determine sample size in clinical trials 4 quantities that must be specified to calculate power: 1. Minimum clinically relevant difference Larger differences will allow for greater separation between populations, so should be easier to detect a difference Smaller differences are harder to detect, will result in greater overlap of populations and thus require greater sample sizes and have an increased type 2 error 2. Standard deviation The standard deviation (σ) of the population from which a data set is sampled affects the probability of error for a given sample size. For tightly distributed data (with a relatively small σ), a statistical test is more likely to detect a significant difference Versus loosely distributed data with a relatively high σ and more potential for overlap of the distributions. When undertaking research, the population σ may not be known. Estimates may be obtained from previously published data or by the conduct of a pilot study. Methods also exist for empirical estimation 3. Type 1 error rate (α = 0.05) 4. Type 2 error rate (β = 0.2)

Answer 40

– Is the data normal (parametric) or not? – Are the results paired or unpaired? – Are the outcomes dichotomous (eg dead/alive)? – Are 2 groups being compared, or more? Distribution tests Smaller sample sizes may not appear normally distributed, even though the data is compatible with a normally distributed population -> therefore, need to perform statistical tests to determine whether data is normal In very small samples sizes (<20), these tests may give false positive; if in doubt, use non parametric statistical analysis, accepting that the analysis may lack power Eg Shapiro–Wilkes test, D'Agostino–Pearson omnibus test Student’s unpaired test is used to compare two groups of normally distributed independent groups Student’s paired t-test is used to compare two groups of normally distributed matched groups Analysis of variance (ANOVA) is used to compare three or more groups of normally distributed independent groups Repeated measures ANOVA is used to compare three or more groups of normally distributed matched groups Categorical data are compared by drawing-up a contingency table and applying either Fisher’s exact or χ2 tests

Answer 41

* Relationships may exist between 2 variables eg: Weight and Propofol requirements * If data sets are normally distributed: Pearson Correlation * If data is not normal: Spearmans rho (ρ) or Kendall’s Tau (τ) * These tests generate a correlation coefficient, stretching between -1 and +1 * The statistical significance (p) of a correlation coefficient can be measured Correlation is misused when comparing the results of 2 diagnostic tests or measurement devices * It should be expected the correlation is extremely high * In this case, the Bland Altman test is used

Answer 42

* Regression is used when a significant relationship exists between 2 data sets * Used to extrapolate and predict unknown values * Linear or non-linear

Answer 43

Probability that an event will occur Only prospective studies can generate relative risk Absolute Risk is the risk of an event occurring in the exposed group. Relative Risk (or risk ratio) is the risk of an event occurring in the exposed group relative to the unexposed group.

Answer 44

Number of times an event happens / number of times an event does not happen Retrospective studies are only capable of generating odds Odds are the probability of an event happening compared to the probability of it not happening, usually expressed as a fraction Coin Flip: Risk of heads = 50% Odds of heads = 1 (50%/50% = 1) The Odds Ratio is the ratio of the odds of the outcome occurring in the exposed compared to the odds of it occurring in the unexposed Odds used when: The denominator is uncertain, i.e.: In retrospective designs, such as case-control studies when patients with the disease were identified, and then exposures ascertained When it statistically appropriate (ORs are much easier to use in statistical tests), i.e.: Multivariate regression Systematic Reviews

Answer 45

risk outcome (control) - risk outcome (study)

Answer 46

1/ absolute risk reduction ARR = risk outcome (control) - risk outcome (study)

Answer 47

An OR < 1 suggests the risk is lower in the exposed group An OR > 1 suggests the risk is higher in the exposed group An OR = 1 suggests that the groups are equivalent In general, the OR overstates risk compared to the Relative Risk It is approximately equal to the RR when the outcome is rare (< 10%) This is because when the event rate is small, the number of non-events in a group is very similar to the overall number of individuals in the group. Larger variance in prevalence = larger difference in RR and OR

Answer 48

True Positive Rate = TP/TP + FN

Answer 49

True Negative Rate = TN/TN + FP

Answer 50

likelihood of true positive

Answer 51

likelihood of true negative

Answer 52

- analytical process and discussion that occurs amongst the team shortly after an event or experience - voluntary process that occurs immediately after the event or experience, or during same shift, preferably while all members of the team are present - should take approximately 5-15 minutes. - focus of a hot debrief is on process and learning rather than outcome. - useful when cognitive load and stress is high as a means to decompress and de-escalate the team before the next theatre case Examples of clinical triggers for hot debriefing: * Exceptional skill and teamwork that you would like to replicate again in the future * Anything that causes global distress or discomfort to the team * Rapid Sequence Induction Intubation (to focus on the positives or process) * High acuity/Low Opportunity Events (HALO)- these are high risk clinical procedures done in life threatening situations to utilise all interventions to rescue a situation that may not be salvageable, E.g. cricothyrotomy. * Events to maximise learning * Events that people may only see once in a career. * Paediatric events, particularly if you are an adult based hospital * Equipment failure * Trauma cases where there has been visually confronting scenes and distressing emotions. * Clinical error or near misses. Examples of non-clinical triggers for hot debriefing: * Following occupational violence or threat of violence * The unexpected death or injury of a staff member outside of work * An internal or external disaster or threat * An incident in the college eg: if all the trainees had to resit their exams due to a technical failure * Any event where there was global staff distress or concerns. Introduction/Intention Reactions Analysis Summary Take homes Thank you

Answer 53

- delayed debrief that occurs days or weeks after a clinical event - to explore in greater depth the clinical event when there is more clinical information\ and more time for the clinicians to invest in the process. - allows time to gain further information and insights into the event and allows the opportunity for all staff who would like to attend to advocate the possibility to do so.

Answer 54

Feedback: giving specific information about a trainee’s observed performance in comparison with an expected standard It is given with the intent to improve the trainee’s performance

Answer 55

Boyer’s Scholarly Domains: 1. discovery, the search for new knowledge 2. integration, bringing findings together from different disciplines or sources 3. application, discovering new ways that knowledge can be used to solve real world problems 4. teaching, applying best practices to develop skills and disseminate knowledge

Answer 56

Glassick’s Criteria of high quality scholarship: 1. clear goals 2. adequate preparation 3. appropriate methods 4. significant results 5. effective presentation 6. reflective critique

Answer 57

1. be original 2. advance the field of health professions education by building on theory, research or best practice 3. be archived and disseminated 4. provide the health professions education community with the ability to comment on and provide feedback in a transparent fashion that informs wider discussion

Answer 58

1. meets the criteria of scholarship outlined by Glassick and Sherbino et al. 2. content is evaluated with respect to its impact, reflecting the extent to which a person’s work reaches its intended audience 3. academics explain their role or ‘brand’ in the social media landscape; helping to establish recognition by others of a scholar’s areas of expertise 4. metrics particular to social media that can help reviewers gauge the overall scientific rigour and quality of a digital scholar’s work

Answer 59

Planning - Define topic to be evaluated - Research -> literature review -> current practice - Type of data to be collected - Method of collection - Analysis of data Data collection & review - Collection & analysis of data - Evaluation of results - Determination of action to be taken Implement changes - Staff education, training, equipment Reaudit - Monitor outcome of introduced changes - Plan repeat audit Policy development - Creation/update guidelines/protocols with improvement Resources - QA coordinator for each anaesthetic department - Sufficient people, time and support should be available for all anaesthetists & trainees to participate

Answer 60

- "one colleague providing a service to another, enabling the provision of better healthcare" - does not examine the patient - has no direct communication with the patient - does not review the patient’s records - has no obligation for formal consultation - receives no payments for services - gives opinion and advice only to the primary treating physician the treating physician remains in control of the patient’s care and treatment

Answer 61

- “a treating doctor asks a patient to see a colleague so that the two doctors can discuss the case with the aim of allowing the treating doctor to better understand the case and its significant aspects. Therefore, a second opinion is provided to and for the treating doctor, to help their assessment and formulation of a treatment plan." 1. members are consulted because of their expertise 2. there may be seen to be a failure to intervene to prevent damage 3. Even if only consulted for a second opinion and not the primary treating practitioner, a duty of care exists

Answer 62

1. patient of interest (or carer) understands the purpose of the MDT meeting 2. patient is aware of the disciplines that may participate 3. patient is informed about those who in the MDT process will be present in an observational capacity 4. patient is informed about what data from their medical history will be shared consideration should be given to adding a fifth condition: patients should be informed about the breadth of their medical history that will be provided and that needs to be considered by the MDT, and why this information is necessary

Answer 63

Equipment — O2 buffer 50-100% — consumables buffer 20% — adequate battery life Monitoring — Pulse rate and ECG trace — Blood pressure (invasive and non-invasive) — Respiratory rate — Oxygen saturation — Continuous waveform capnography for all ventilated and/or sedated patients — Glasgow coma score — Pupil response — Temperature — Ventilator parameters including low/high pressure alarms — Capillary refill IHT/pre hospital: — Blood glucose — Urine output — Pain score — Point of care ultrasound — Point of care blood analysis — Ventilator parameters including low/high pressure alarms — Other compartmental pressures when indicated (e.g., Intracranial Pressure, Intra-abdominal etc.) — Pressure areas

Answer 64

A forest plot is a graph used in meta-analyses (systematic reviews) to visually combine results from multiple studies on the same topic, showing each study's effect size (like odds ratio) as a square, its confidence interval (CI) as whiskers, and a diamond at the bottom for the overall combined result, helping assess treatment impact, study consistency, and statistical significance against a null line (usually at 1 or 0)

Answer 65

1. Avoid routinely performing preoperative blood investigations, chest X-ray or spirometry prior to surgery, but instead order in response to patient factors, symptoms and signs, disease, or planned surgery. 2. Avoid ordering cardiac stress testing for asymptomatic patients prior to undergoing low to intermediate risk non-cardiac surgery. 3. Avoid administering packed red blood cells (blood transfusion) to a young healthy patient with a haemoglobin of ≥70g/L who does not have on-going blood loss, unless the patient is symptomatic or hemodynamically unstable. 4. Avoid initiating anaesthesia for patients with limited life expectancy, at high risk of death or severely impaired functional recovery, without discussing expected outcomes and goals of care. 5. Avoid initiating anaesthesia for patients with significant co-morbidities without adequate, timely preoperative assessment and postoperative facilities to meet their needs. 6. Avoid routine prescription of SR opioids for acute pain unless demonstrated need, close follow up & cessation plan.

Answer 66

Ask: I am concerned about [x], are you OK to talk about it? Assess: How do they feel about [x] Advise: Benefits of treatment/lifestyle changes Assist: Management plan Arrange: Referrals

Answer 67

Consent/Competence/Capacity/Autonomy for Labour Epidural ○ Severe pain may impair or improve informed consent § Severe pain may impair ability/remove capacity to process information on risks and benefits § Vs only with the knowledge of severity of pain can the true risks and benefits be considered § Discussion in clinic does not provide all of the information necessary to make an informed decision as the level of pain is not known ○ Some studies say epidural protects against post-natal depression, others say that the unintended decision to have an epidural increase PND rates ○ Attendance of anaesthetist >15 minutes after request associated with reduced satisfaction ○ Paternalistic withholding justified in the presence of severe coagulopathy due to non-maleficence § The normal risks are not reason enough to deny the request for an epidural We treat the severe acute pain of labour differently to other types and that we tolerate delays and under-resourcing The 30 minute target should be worst-case scenario, not the accepted standard time

Answer 68

ALARA principle – as low as reasonably achievable – is the guiding principle of radiation safety - Time ○ reducing the total time the radiation source is active ○ reducing the time spent near the active source - Distance ○ inverse square law, implies that doubling the distance from the radiation source will effectively reduce exposure four-fold ○ recommendation of ARPANSA to stand at least two metres away from the X-ray tube, and also outside the primary beam - Protection ○ mobile lead shields and lead drapes may not protect from scatter ○ PPE: aprons, thyroid shields, leaded glasses, and gloves ○ minimum of 0.25 mm lead thickness, 0.5mm if high exposure Education, audit, signage, monitoring (dosimeter)

Answer 69

formal RCA process - assembling an independent multidisciplinary team - reconstructing the event through documentation and interviews - identifying systemic factors - formulating achievable recommendations * structured process * focus on system-level factors rather than individual blame * RCA contributes to quality improvement by providing actionable recommendations and monitoring the effectiveness of implemented changes

Answer 70

- Perform literature review - Formulate question (PICO) and hypothesis - Gather resources: research staff, stakeholders, funding, data infrastructure - Draft study protocol & study materials (consent form, data collection form) - Ethics and governance approval - Consider pilot study - Assists to confirm power calculation / outcome event rates - recruitment rates - safety endpoints - feasibility - Re-evaluate pilot data / adjust protocol as needed - Rollout full-scale study - Recruitment / data collection - Analysis - Disseminate research (manuscript / presentation etc)

Answer 71

Meta-analyses are considered Level 1 evidence Pool data across multiple studies, with weighting according to sample size, variance, and statistical tests for agreement with other studies (heterogeneity - low if all studies are either positive / negative, high if they all disagree, generate statistics) Advantage: - Increased statistical power through larger number of patients -> more precise estimate - Forrest plot generated - Objective appraisal - Reduce likelihood of false negative - Heterogeneity between studies, including biases, may average out -> hence if a signal is found, it is more likely to be true Disadvantage: - Process of including studies for meta-analysis may be biased - Ability to pool data depends on - Accessibility of original data - some researchers may not disclose - Quality of the original data - Heterogeneity - Bias - Publication bias - negative studies are often not published -> treatment effect may be unduly skewed towards positive - Funnel plot used to assess publication bias - Blur the context by which the individual data is generated: - Summary data used instead of individual data - Inclusion / exclusion may not be detailed - Hence, pooled data may not be externally valid

Answer 72

Multicentre RCTs - Useful in evaluating a treatment compared to a control - Observational study has no control - ie. Simply observe outcome overtime, sometime in response to an exposure (eg. GA vs spinal) but not designated allocation to group Advantages: - Robust designs - Level 2 evidence - considered moderate - high certainty evidence in NHMRC levels of evidence - Compared with observation study - low-moderate certainty as Level 3 - Randomisation eliminate potential confounders - Blinding will further enhance internal validity by minimising observer's bias - Observational study will have more bias due to lack of randomisation, blinding - Lack of control group will also limit ability to make statistical inference - Large number of patients = more precise estimate of treatment effect - Greater power = reduces probability of type 2 error - Single institution study will naturally have lower number of patient - results more prone to selection bias - Reduces confounding effect of regional differences in practice, population, disease prevalence - Enhances external validity, whereas finding of an observational study may only be relevant to the context of that particular institution, hence must be interpreted with care Disadvantage: - Higher cost to conduct - need funding at national or multinational level - Financial interest may introduce bias - More time consuming to plan, gain ethical approval, execute and report not appropriate for QA & QI purpose where as observational study is relatively quick and hence yielding valuable information - (Risk of intervention to participants) - Observational study = no risk or very low risk - RCTs = may be moderate-high In addition, observational study may be hypothesis generating - useful for further research which may impact clinical practice

Answer 73

RANDOM error (chance / imprecision) - Error arising from natural variability because a sample, not the whole population, is studied - Affects precision, not validity - Reduces with larger sample size Sources: Sampling variability - Small samples → unstable estimates, wide confidence intervals Type I error (α error) - False positive: rejecting a true null hypothesis. - Governed by the clinical significance of α - Conventionally α = 0.05. - Common causes: Multiple hypothesis testing, Small sample with chance extreme results, Data dredging / p-hacking, Flexible outcome definitions post-hoc, Early stopping without proper statistical correction Type II error (β error) - False negative: failing to detect a true effect. - More common than type 1 error - Related to low statistical power (Power = 1 − β). - Common causes: Underpowered studies, Small sample size, Overestimated effect size in planning, High outcome variability, Poor measurement precision, High loss to follow-up Multiple comparisons - Repeated hypothesis testing inflates Type I error if not adjusted. Random measurement error - Imprecise instruments or inconsistent measurement increases variability SYSTEMIC error (bias) - Error that consistently distorts the estimate in one direction. - Affects validity (accuracy). - Does NOT improve with increasing sample size. - Not quantifiable with p-values or confidence intervals - Leads to consistently incorrect conclusions Sources: Systematic measurement error - Faulty calibration, consistent observer bias. - Example: BP cuff reads 10 mmHg too high. Differential measurement error - Measurement differs between comparison groups → biased effect estimate. Misclassification error - Incorrect categorisation of exposure or outcome Selection bias - Study population differs systematically from target population. Attrition bias - Differential loss to follow-up between groups. Allocation bias - Predictable or unconcealed treatment assignment in trials. Confounding - A third variable: - Associated with the exposure - Independently affects the outcome - Not on the causal pathway - Creates a false association or masks a real one. Information bias - Recall bias: Differential memory of past exposure. - Observer/detection bias: Outcome assessment influenced by knowledge of exposure. - Reporting bias: Selective outcome reporting or p-hacking. Missing data error - Missing not at random (MNAR) - Probability related to outcome or exposure. - Can significantly bias results. Model and analysis-related error - Incorrect statistical model: Violated assumptions (normality, independence). - Omitted variables: Failure to include important confounders. - Overfitting: Too many predictors for sample size. - Data dredging: Hypotheses generated after looking at results.

Answer 74

one subject undergoes both intervention (acts as own control) Crossover study: all subjects recruited receive either treatment A or treatment B (the order decided by random allocation for each patient), followed by the other treatment Paired statistical tests are more powerful and fewer subjects need to be recruited in order to prove a given difference between the study groups alternative: Matching = Prospective matching patients for important characteristics (potentially unreliable) - twin studies best version of this

Answer 75

- The structure of anaesthesia and perioperative care must support safe, high quality care. - The provision of anaesthesia is a medical role. - Innovation should be based on delegation, not substitution, of roles. - Anaesthesia and perioperative care require a team of highly skilled health professionals. - New or extended roles within the anaesthesia and perioperative care team should be developed, implemented and evaluated in a systematic and consistent way.

Professionalism, stats Flashcards

(108 cards)