Ask
| Site: | Learn: free, high quality, CPD for the veterinary profession |
| Course: | EBVM Learning |
| Book: | Ask |
| Printed by: | Guest user |
| Date: | Wednesday, 1 May 2024, 10:46 PM |
1. Introduction
1. Introduction
The first step in practising EBVM is to ask the 'right' question(s). Without the right question, we cannot search and Acquire the correct evidence for critical appraisal, nor can we establish a context within which we can Appraise its relevance and quality. Only then can we Apply our new knowledge in a clinical context, in order to Assess its impact on our practice.
By the end of this section you will be able to:
- describe why a well-formed question is fundamental to the EBVM process and avoid the common pitfalls in asking questions
- identify clinical questions in practice
- use the (S)PICO mnemonic to construct a searchable clinical question.
2. The importance of starting with a good question
2. The importance of starting with a good question
Questioning our current practice underpins the principles of EBVM – in order to practice EBVM, we must be prepared to question what we do and change accordingly.
By questioning our practice in a critical way, we can move in a direction that keeps us up to date; also, by using the best possible evidence, we can offer our patients the best possible outcomes.
Well-formed questions underpin the very core of scientific methodology:
One cannot get a clear answer to a vague question. The language of science is particularly distinguished by the fact it centres around well-stated questions. (Johnson, 1946)
One of the most common mistakes those new to EBVM might make is to start searching for answers with only a vague idea of what information is needed. To address complex or poorly defined clinical problems, you must first break these problems down into a series of more precise questions. By framing your questions in this narrow, precise way, you increase your likelihood of finding evidence that specifically answers your question. The process of formulating these precise questions will focus your thoughts on the problem you are addressing, your clinical choices and outcome values.
TIP: In equine medicine, rather than asking “What should I do about recurrent laryngeal neuropathy in horses I see in my practice?“, you could ask “In adult, racing thoroughbred horses presenting with recurrent laryngeal neuropathy, does ventriculectomy (‘Hobday’) with ventriculocordectomy, compared with prosthetic laryngoplasty (‘tie-back’), have a greater success rate for return to racing?”
This question could also have many variations: for instance, you could change the outcome you want to measure, and ask “In adult, racing thoroughbred horses presenting with recurrent laryngeal neuropathy, does ventriculectomy (‘Hobday’) with ventriculocordectomy, compared with prosthetic laryngoplasty (‘tie-back’), have a greater reduction in air turbulence?”
3. Types of clinical questions
3. Types of clinical questions
To benefit both patients and clinicians, questions need to be focused and directly relevant to the patient or scenario at hand.
Categorising the type of your clinical
question can help you to decide which study design would best answer your question and its level of evidence, which becomes important when you begin to appraise the evidence.
In the Appraise section, we will explore the different study types and the levels of evidence in more detail.
Clinical questions can be divided into five main topic areas, relating to:
- treatment
- prognosis and incidence
- aetiology or risk
- diagnosis
- prevalence.
Treatment
These types of questions refer to treatment choices made about patients in practice. These choices can include drugs or medicines to be used, surgical methods, changes in diet or management, and many more. These types of questions are best answered by randomised controlled trials when they are available.
Example: Which diet is best to feed cats with chronic renal disease?
Prognosis and incidence
These types of questions relate to the likelihood of disease or the progression of disease over time. These questions are best answered by cohort studies.
Example: Does sex affect survival in flat-coat retrievers with cancer?
Aetiology and risk
These types of questions investigate the origin of disease or the factors influencing development of a certain condition or disease. These questions are best answered by cohort studies, case-control studies or cross-sectional studies.
Example: What are the risks of adverse events in general anaesthesia in ferrets under different protocols?
Diagnosis
These types of questions involve identification of a disorder based on the animal’s presenting signs. These questions are best answered by diagnostic test validation studies (also known as diagnostic evaluation studies).
Example: Which diagnostic test is most reliable for diagnosing fascioliasis in dairy cattle?
Prevalence
These questions consider the frequency of disease at a certain point in time and are best answered by cross-sectional studies.
Example: What is the prevalence of cardiac disorders in Welsh Section A mountain ponies?
Stakeholder experiences, preferences and values
We can also ask questions about the experiences, values and preferences of the stakeholders concerned, which, while not necessarily clinical questions, are relevant to the more holistic practice of EBVM.
Stakeholder experiences, preferences and values questions consider a wide range of issues, and no one specific type of study is sufficient to address this general category. Both qualitative, quantitative and mixed method approaches drawn from both social and biological science modalities may be appropriate, and a full guide to this is currently beyond the remit of this course.
4. How to construct a good EBVM question?
4. How to construct a good EBVM question?
A well-formed clinical question is the most efficient route to obtaining a clear answer to the problem or challenge you are interested in. The question you ask needs to be formatted in such a way as to aid you in your search for answers.
Formatting your question correctly is important in ensuring that your search for evidence is structured, systematic and complete. See the Acquire section of this resource for more details.
Various systems have been developed to assist practitioners in formatting their clinical problems into useful questions, enabling a structured, systematic and complete search of the evidence. The system depends on the type of question being asked.
The most common system used to format a question is the PICO system, focusing on the:
- P – Patient: population and/or problem
- I – Intervention: treatment, or thing of Interest: prognostic factor or exposure
- C – Comparator: comparison or control
- O – Outcome
We will focus on the PICO system in this resource. Sometimes it is adapted to (S)PICO where the ‘S’ stands for species. You will commonly see PICO used and sometimes SPICO. Species is part of the patient definition (in ‘P’) but adding the ‘S’ will ensure you don’t forget it.
4.1. (S) P – Species, Patient: population and/or problem
4.1 (S) P – Species, Patient: population and/or problem
Species - that bit is easy! The next step in formulating a clinical question in the (S)PICO format is to consider the patient and the clinical problem you are faced with.
It is helpful to think in terms of the population you are dealing with and to characterise your patient in general terms (e.g. a geriatric cat with a diagnosis of chronic kidney disease).
Table 1: Examples of questions expressed as species, patient and clinical problem
| QUESTION | SPECIES/PATIENT/PROBLEM |
|---|---|
|
Which diet is best to feed to cats with chronic renal disease? |
Cats with chronic renal disease |
|
Which diagnostic test is most reliable for diagnosing fascioliasis in lactating dairy cattle? |
Diagnosis of fascioliasis in lactating dairy cattle |
|
Does sex affect survival in flat-coat retrievers with cutaneous lymphoma? |
Flat-coated retrievers with cutaneous lymphoma |
|
Are the risks of inhalational induction of general anaesthesia higher compared to injectable induction ferrets under different protocols? |
Ferrets undergoing general anaesthesia |
|
Are cardiac disorders in Welsh Section A mountain ponies more prevalent than other breeds? |
Horses |
4.2. I – Intervention: treatment, prognostic factor or exposure
4.2 I – Intervention: treatment, prognostic factor or exposure
You might be interested in a specific treatment, a factor that will indicate prognosis in a disease process, or the association of a certain exposure with disease, depending on the question.
These interventions are often considered with their matching comparators – something you might compare against the group receiving the intervention (see the next page for further information about comparators).
Table 2: Examples of questions expressed as intervention or interest
| QUESTION | INTERVENTION/INTEREST |
|---|---|
|
Which diet is best to feed cats with chronic renal disease? |
Feeding a renal prescription diet |
|
Which diagnostic test is most reliable for diagnosing fascioliasis in lactating dairy cattle? |
Milk ELISA |
|
Does sex affect survival in flat-coat retrievers with cutaneous lymphoma? |
Being male |
|
Are the risks of inhalational induction of general anaesthesia higher compared to injectable induction ferrets under different protocols? |
General anaesthesia induction by triple injectable agent |
|
Are cardiac disorders in Welsh Section A mountain ponies more prevalent than other breeds? |
Being a Welsh Section A mountain pony |
4.3. C – Comparator: comparison or control
4.3 C – Comparator: comparison or control
Now that you have defined your population and intervention of interest, you need to consider your choices (i.e. what the intervention will be compared to).
It is important to realise that any intervention needs to be considered at the same time as a comparator, as without a comparison it is difficult to evaluate the impact of the particular treatment, prognostic factor or exposure you are interested in.
Table 3: Examples of questions with comparators
| QUESTION | COMPARATOR |
|---|---|
|
Which diet is best to feed cats with chronic renal disease? |
Not feeding a renal prescription diet |
|
Which diagnostic test is most reliable for diagnosing fascioliasis in lactating dairy cattle? |
Serum ELISA |
|
Does sex affect survival in flat-coat retrievers with cutaneous lymphoma? |
Being female |
|
Are the risks of inhalational induction of general anaesthesia higher compared to injectable induction ferrets under different protocols? |
General anaesthesia induction by inhalational agent |
|
Are cardiac disorders in Welsh Section A mountain ponies more prevalent than other breeds? |
Being any other breed of horse |
4.4. O – Outcome by which 'I' will be compared with 'C'
4.4 O – Outcome by which 'I' will be compared with 'C'
Choosing a specific desired outcome is a key part of formulating an evidence-based question about a patient.
This ensures you will Acquire, Appraise and Apply evidence pertaining to the specific outcome of interest for you and your individual patient.
Table 4: Examples of questions and outcomes
| QUESTION | OUTCOME |
|---|---|
|
Which diet is best to feed cats with chronic renal disease? |
Survival time |
|
Which diagnostic test is most reliable for diagnosing fascioliasis in lactating dairy cattle? |
Predictive values |
|
Does sex affect survival in flat-coat retrievers with cutaneous lymphoma? |
Average life expectancy |
|
Are the risks of inhalational induction of general anaesthesia higher compared to injectable induction ferrets under different protocols? |
Mortality rate |
|
Are cardiac disorders in Welsh Section A mountain ponies more prevalent than other breeds? |
Prevalence of cardiac disorders |
4.5. Example (S)PICO questions
4.5 Example (S)PICO questions
You have now looked at all the elements required to construct a full (S)PICO question.
Below are structured example (S)PICO questions created around the examples from the previous section ‘How to construct a good EBVM question?’ for the five different types of clinical question.
Original question: Which diet is best to feed cats with chronic renal disease?
(S)PICO: In [cats with chronic renal disease] does [feeding a renal prescription diet] compared with [not feeding a renal prescription diet] impact on [survival time]?
Original question: Which diagnostic test is most reliable for diagnosing fascioliasis in dairy cattle?
(S)PICO: In [lactating dairy cattle] does the [milk ELISA] compared with [serum ELISA] have better [positive and negative predictive values] for [diagnosing fascioliasis]?
Original question: Does sex affect survival in flat-coated retrievers with cutaneous lymphoma?
(S)PICO: In [flat-coated retrievers with cutaneous lymphoma], does [being a male] compared with [being a female] affect [average life expectancy]?
Original question: What are the risks of inducing general anaesthesia in ferrets under different protocols?
(S)PICO: In [ferrets undergoing general anaesthesia], what is the [risk of death] under general anaesthesia induced by [triple injectable agent] compared with the [inhalational agent]?
Original question: What is the prevalence of cardiac disorders in Welsh Section A mountain ponies?
(S)PICO: In [horses], does [being a Welsh Section A mountain pony] compared with [being any other breed] increase the [prevalence of cardiac disorders]?
The (S)PICO framework can be applied to most clinical questions and is easy to use once you have learned its salient principles.
5. Challenges to building a PICO
5. Challenges to building a PICO
Once you have identified your clinical question and created your PICO, there are a few more considerations to inform and re-adjust your PICO and get the results you need.
Scope versus volume of evidence
The Acquire section, coming next, explains more fully how you can use your PICO question to find the evidence.
Sometimes the question you ask may yield too much or too little relevant evidence. For example, you may find that your PICO has only yielded two papers, neither of which entirely answers your question. Assuming your search was conducted thoroughly, this may mean there is not enough evidence available to answer your question.
On the other hand, you may find that your search yields dozens and dozens of results, not all of which specifically relate to the problem or question you have in mind. In this case, it might be necessary to adjust your PICO to be narrower and more focussed in order to find only the most relevant evidence to answer your question. How you implement the evidence into practice will be further covered in Apply and Assess.
Choice of interventions and comparators
To cover all interventions, interests and comparators, multiple PICO questions need to be formed (often easier!) or you can choose a more general question. If you choose the latter, then you may end up with more evidence to sift through and the search outputs may become less relevant.
Narrow question: In [dogs with osteoarthritis] is [meloxicam] better than [tramadol] at [reducing pain]?
Wider question: In [dogs with osteoarthritis] are [NSAIDs] better than [tramadol] at [reducing pain]?
Multiple outcomes
Sometimes, a clear choice for your patient will only have a single objective desirable outcome, and it is certainly nice when this is the case in your clinical question. In reality, however, veterinary professionals often want to investigate a number of different outcomes for ourselves, our clients and our patients. For example, we may wish for a treatment that is effective, safe, easy to use and economic.
Many studies in the literature may address multiple outcomes, looking at both efficacy (e.g. survival times) and negative outcomes (e.g. adverse events) as well as costs, all in the same study. Sometimes you may need to look across multiple studies to gather these data; to do this, you will effectively be asking a series of PICO questions, all with different outcomes. One strategy is to refine your outcome to be a composite statement that reflects your overall aims for a case (e.g. ‘long-term survival whilst pain free’).
6. Example scenarios using the PICO format
6. Example scenarios using the PICO format
A series of example case scenarios for you to consider are now given.
For each example, we suggest you attempt to write out a PICO question, and then expand the text to see an example provided in the PICO format.
You can use the tool PICO.vet to help you build a well-structured and focused clinical question.
Clinical Scenario
Carprofen and local anaesthesia in calves undergoing disbudding
During a visit to one of your small beef herds, the owner, Mary Reader, asks you to disbud three calves that have been born within the last few weeks. The last time Mrs Reader had animals disbudded with hot iron cautery, she was upset by how the animals behaved after the procedure – she says they were “out of sorts” and looked uncomfortable.
Whilst Mrs Reader knows the animals must be disbudded, she asks you whether analgesics as well as the local anaesthetic would be likely to reduce the pain from the disbudding. You assure her you always use local anaesthetic, but you wonder if the addition of carprofen would decrease the level of discomfort experienced by the calves.
In [calves undergoing non-chemical disbudding] does [the administration of carprofen in addition to local anaesthetic] versus [local anaesthetic alone] [decrease the behavioural indicators of discomfort associated with the procedure]?
Clinical Scenario
Carprofen in dairy cattle with toxic mastitis
During a visit to one of your
dairy farms, the owner Steve Jones comments that he has had a number of cows sick with mastitis, which he thinks are caused by E. coli. He finds these cases very difficult to treat.
Steve has recently seen an advert for an anti-inflammatory drug containing carprofen, which claims that it will improve the recovery rate of cows with toxic mastitis. Over the years, you have had a number of conversations with Steve about the relative merits of different treatment options for cases of toxic E. coli mastitis. You wonder if carprofen would make a difference to recovery of cows on Steve’s farm.
In [dairy cattle with E. coli mastitis] does [the administration of carprofen] compared to [no anti-inflammatory treatment] [improve clinical recovery]?
Clinical Scenario
Anchoring versus pocket technique for surgical repair of cherry eye in dogs
You are presented with a 1-year-old Beagle with a unilateral cherry eye. It has been present for two months and is not bothering the dog. The owner wants to know what to do. You ring the two veterinary ophthalmologists in the local area for advice. One routinely performs an anchoring technique, whilst the other has had good results with a mucosal pocket technique. Having only two expert opinions to go by, you decide to look for any available higher-level evidence.
In [dogs undergoing surgery for repair of a prolapsed gland of the third eyelid (cherry eye)], is [a pocket technique] superior to [an anchoring technique] in [preventing recurrence]?
Clinical Scenario
Renal diets in cats with chronic kidney disease
Chloe, a 14-year-old
domestic shorthaired cat, has just been diagnosed with IRIS late stage II kidney disease. She is not proteinuric, and her blood pressure is normal. You have stabilised her azotaemia, and her appetite is now good. What is the benefit of
a kidney prescription diet for this cat?
In [cats with naturally occurring chronic kidney disease] does [a renal prescription diet] compared to [normal diet] increase the [survival time] of affected cats?
8. Summary
8. Summary
Learning outcomes
You should now be more familiar with how to:
- describe why a well-formed question is fundamental to the EBVM process and avoid the common pitfalls in asking questions
- identify clinical questions in practice
- use the (S)PICO mnemonic to construct a searchable clinical question.
9. References
9. References
Cockcroft, P. D. and Holmes, M. A. (2003) Handbook of Evidence-based Veterinary Medicine. Oxford: Blackwell Publishing
Johnson, W. (1946) People in quandaries: the semantics of personal adjustment.New York: Harper & Row
Lanyon, L. (2014) Evidence-based veterinary medicine: a clear and present challenge. Veterinary Record, 174 (7), pp. 173-175