Am I muddled? Figure 3 is labelled as an example of 'high specificity'. Just following the simple rules set out for figure 2 shows that: specificity = number of true negatives / (true negatives + false positives) = 8/ (8 + 62) = 0.1111 or 11% & sensitivity = number of true positives (true positives + false negatives) = 30 (30 + 0) = 1 or 100%. Surely this is a highly sensitive test??
Comment from the Editor: Thanks to Dr. McDermott for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1 We hope it addresses all your concerns.
I may have missed something but the Figure 3 does not illustrate high specificity, it shows perfect sensitivity and very low specificity. There is the same confusion in the paragraph that explains it "There is a risk that a test with high specificity will capture some people who do not have Disease D (figure 3). The screening test in figure 2 will capture all those who have the disease but also many who do not. " That would be a test with high sensitivity. And the reference to the figures is confusing (I believe the authors are referring to Figure 3 all along).
"Comment from the Editor: Thanks to Dr. Lassale for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1"
I read this article with interest as I revised for an recent assessment. I have some comments which may be of use to you and your readers, as I feel there are mistakes in the use of the terms sensitivity and specificity, which seem to be interchanged at various points throughout the article, and notably in the second paragraph, where there are initially defined. From my understanding, the ability of a test to correctly identify a disease is sensitivity, and the ability to correctly identify the absence of a disease is specificity, rather than he reverse as presented here. This may be typographical, as the authors go on to correctly illustrate both sensitivity and specificity in the subsequent example, and in Box 1, however further instances of this error appear later.
Another prominent example is Fig 3 and the associated paragraph, which I believe refers to high sensitivity, rather than specificity. As specificity is TN/(TN+FP), a higher false positive rate would decrease specificity. The figure shows a test which is highly sensitive ((TP/(TP+FN) = 30/(30+0) = 1 = 100%), but with low specificity ((TN/(TN+FP) = 8/(8+62) = 8/70 = 0.11 = 11%).
There seems to be a further confusion, particularly in the paragraph using the Ottawa ankle rule as an illustrative example. The 'wide range of sensitivity' discussed early in this paragraph, is not sensitivity, but specificity as stated in the discussion of the cited article (8). Indeed, this is indicated in t...
I read this article with interest as I revised for an recent assessment. I have some comments which may be of use to you and your readers, as I feel there are mistakes in the use of the terms sensitivity and specificity, which seem to be interchanged at various points throughout the article, and notably in the second paragraph, where there are initially defined. From my understanding, the ability of a test to correctly identify a disease is sensitivity, and the ability to correctly identify the absence of a disease is specificity, rather than he reverse as presented here. This may be typographical, as the authors go on to correctly illustrate both sensitivity and specificity in the subsequent example, and in Box 1, however further instances of this error appear later.
Another prominent example is Fig 3 and the associated paragraph, which I believe refers to high sensitivity, rather than specificity. As specificity is TN/(TN+FP), a higher false positive rate would decrease specificity. The figure shows a test which is highly sensitive ((TP/(TP+FN) = 30/(30+0) = 1 = 100%), but with low specificity ((TN/(TN+FP) = 8/(8+62) = 8/70 = 0.11 = 11%).
There seems to be a further confusion, particularly in the paragraph using the Ottawa ankle rule as an illustrative example. The 'wide range of sensitivity' discussed early in this paragraph, is not sensitivity, but specificity as stated in the discussion of the cited article (8). Indeed, this is indicated in the paragraph by the assertion of the test ability to 'rule out' ankle fracture, which is generally applied to the specificity of a test.
The terms sensitivity and specificity again seem to be erroneously used, or interchanged, at various points throughout this illustrative paragraph - for example 'A specificity of 96% means that when the rules are applied almost everyone who has a fracture will be selected for an x-ray' is describing sensitivity, rather than specificity - a 96% specificity would indicate that 96% of people not selected for an x-ray by screening would not have a fracture.
The following worked example then alternates between using 95% and 96% specificity, and while this is likely typographical, the terms sensitivity and specificity seem to again be interchanged. The example asks the reader to assume a specificity of 95% (or 96%), which is then substituted into an equation as the result of TP/(TP+FN), which is the equation for sensitivity, not specificity. The inverse is true of the next equation, illustrating sensitivity.
I do not wish to be unduly critical, but as a primer for sensitivity and specificity, it seems an important oversight that the key terms of sensitivity and specificity seem to be exchanged throughout. Of course, I am more than happy to be corrected, and if so would consider it valuable for my own education.
Comment from the Editor: Thanks to Dr. Codd for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1
I think these have been confused in the "definitions"
"Comment from the Editor: Thanks to Dr. Nicholl for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1"
I am writing to highlight what I believe to be several fundamental errors in this article in trying to explain sensitivity and specificity, titled “What are sensitivity and specificity?”
1) Partway through, the article states, "there is a risk that a test with high specificity will capture some people who do not have Disease D (figure 3)."
I believe this should state "high sensitivity", not high specificity, as this would be the complete opposite with low false positives. The same applies for the "high specificity" label for Figure 3 - this illustrates a high sensitivity test (with very low specificity as it gives many false positives), not a high specificity test.
2) In reference to the Ottawa ankle rules:
"They have been shown (in a systematic review) to correctly identify approximately 96% of people who have a fracture and to correctly rule out between 10% and 70% of those who do not have a fracture.(8) The wide range of sensitivity is likely to be due to differences in the education of the clinicians involved".
Assuming a positive test here identifies ankle fractures, I believe this is a wide range in 'specificity', not a "wide range of sensitivity" (which appears to be c.96%).
Sensitivity and specificity are repeatedly swapped incorrectly in this sect...
I am writing to highlight what I believe to be several fundamental errors in this article in trying to explain sensitivity and specificity, titled “What are sensitivity and specificity?”
1) Partway through, the article states, "there is a risk that a test with high specificity will capture some people who do not have Disease D (figure 3)."
I believe this should state "high sensitivity", not high specificity, as this would be the complete opposite with low false positives. The same applies for the "high specificity" label for Figure 3 - this illustrates a high sensitivity test (with very low specificity as it gives many false positives), not a high specificity test.
2) In reference to the Ottawa ankle rules:
"They have been shown (in a systematic review) to correctly identify approximately 96% of people who have a fracture and to correctly rule out between 10% and 70% of those who do not have a fracture.(8) The wide range of sensitivity is likely to be due to differences in the education of the clinicians involved".
Assuming a positive test here identifies ankle fractures, I believe this is a wide range in 'specificity', not a "wide range of sensitivity" (which appears to be c.96%).
Sensitivity and specificity are repeatedly swapped incorrectly in this section. For example, with “A specificity of 96% means that when the rules are applied almost everyone who has a fracture will be selected for an x-ray”. Again, this is a ‘sensitivity’ of 96%.
3) All three uses of the specificity and sensitivity formulas for the Ottawa ankle rules are wrong. The authors input the prevalence (15%) in the numerator for sensitivity, rather than the true positive rate; and use (1 - 15%) in the numerator for specificity instead of the true negative rate. This is followed by the absurd result that “the equation for the lower specificity (0.1=85/(85+x) =765) shows that up to 765 might be sent for an unnecessary x-ray” out of every 100 patients. (This is more false positives than patients.)
The correct rates are:
- False negative probability given the patient has a fracture = (1 – sensitivity) = (1 – 96%) = 4%. (The authors wish to convert this to a percentage of total patients, which would be 4% * 15% prevalence = 0.60 in every 100 patients, not 0.63% using the formula incorrectly)
- Assuming 10% Ottawa test specificity, this implies 90% of the 85 patients without fractures are wrongly suspected of having fractures (false positives). So, 90% * 85 = 76.5 out of 100 patients are incorrectly sent to x-ray
- Assuming 70% Ottawa test specificity, this implies 30% of the 85 patients without fractures are wrongly suspected of having fractures (false positives). So, 30% * 85 = 25.5 out of 100 patients are incorrectly sent to x-ray (not 36 as stated in the article)
This is not an exhaustive list of errors.
I am not a medical professional. However, I expect many professionals rely on the BMJ for accurate information (especially in relation to testing sensitivity and specificity in the current global pandemic), and this inaccurate article has now been viewed in full over 40,000 times and shared widely online. Please correct this article as soon as possible.
"Comment from the Editor: Thanks to Dr. Yash for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1"
This reply is in response to the above published article on March 22, 2022. The article was very informative and captivating due to the focus being on relations to "The impact COVID-19 on nursing students". After reading this article I was able to understand the prevalence of mental health problems and sleep disturbance among nursing students during the pandemic. The pandemic has brought so much hurt to persons between losing a family or losing a job or anything beneficial to help them provide. The effect the pandemic has on students, has been very disturbing. As the article spoke about how transitioning has affected the mental health and sleeping problems, it was so relatable because in The Bahamas, student nurses and registered nurses have been affected drastically.
This article has given me a more global view on how COVID-19 made a negative impact. In the Bahamas, COVID- 19 devastated all forms of jobs and the people that make up the jobs. The results from the survey from the students inside this article showed how 80% were worried about the impact of transition and 62% were worried about management of the workload. In The Bahamas, nursing students were affected in several ways, such as not being able to find resources to do the classes, not being able to complete any labs or clinicals which became a set back and placed myself and others in a very depressing state. In the article under “As a nurse” due to the need of nurses, students were pushed at the f...
This reply is in response to the above published article on March 22, 2022. The article was very informative and captivating due to the focus being on relations to "The impact COVID-19 on nursing students". After reading this article I was able to understand the prevalence of mental health problems and sleep disturbance among nursing students during the pandemic. The pandemic has brought so much hurt to persons between losing a family or losing a job or anything beneficial to help them provide. The effect the pandemic has on students, has been very disturbing. As the article spoke about how transitioning has affected the mental health and sleeping problems, it was so relatable because in The Bahamas, student nurses and registered nurses have been affected drastically.
This article has given me a more global view on how COVID-19 made a negative impact. In the Bahamas, COVID- 19 devastated all forms of jobs and the people that make up the jobs. The results from the survey from the students inside this article showed how 80% were worried about the impact of transition and 62% were worried about management of the workload. In The Bahamas, nursing students were affected in several ways, such as not being able to find resources to do the classes, not being able to complete any labs or clinicals which became a set back and placed myself and others in a very depressing state. In the article under “As a nurse” due to the need of nurses, students were pushed at the frontline with the professionals, and in my opinion, it was not fair, it may have been a good experience or a kind gesture but, the students were basically rushed to perform duties they may have not been ready for.
As I conclude, being a family person will have an affect on a person in the healthcare system especially because they work amongst the disease that’s travelling. Working in a pandemic will bother a person that has a family to go home to because, the thought of being a causative factor as to why someone who is so close to you caught the disease can mentally and emotionally influence the individual. In this article it spoke about how affecting the ones you love can take a toll on a person especially if they are a child or an elderly person. COVID-19 has impacted nurses as student nurses, registered nurses, and the people around them. This article was very relatable because the same things are happening in The Bahamas, this article would be a good source for students writing papers about how COVID-19 has affected the aspect of becoming a nurse and it also helps people to know what is going on in the world just not where they are.
Response to: Dos Santos WM Use of personal protective equipment reduces the risk of contamination by highly infectious diseases such as COVID-19 Evidence-Based Nursing 2021: 24:41
Nurses need accessible and thorough evidence-based guidance to support safe, effective care due to the risk of SARS-CoV-2/COVID-19 infection, sickness, and death. Selection and use of appropriate Personal Protective Equipment (PPE) is fundamental to respiratory protection programs and access to the latest evidence is vital to underpin practice and policy decisions.
The Evidence-Based Nursing commentary by Dr Dos Santos on a recent Cochrane Review of PPE use in healthcare workers offers some useful insights. However, there are other key details and findings from the original review that are also important for readers. While covering more of the body appeared to result in enhanced protection, more difficult donning, doffing, and poorer user comfort could lead to higher contamination risk. Also, more breathable PPE might result in a similar number of contamination spots on the user’s trunk compared to less permeable materials (MD 1.60, 95% CI −0.15 to 3.35) and potentially increase user satisfaction (MD −0.46, 95% CI −0.84 to −0.08).
Centers for Disease Control and Prevention guidance resulted in less contamination compared with using no guidance (small patches: MD −5.44, 95% CI −7.43 to −3.45). Other key findings include:
- One-step glove plus gown removal versus separa...
Response to: Dos Santos WM Use of personal protective equipment reduces the risk of contamination by highly infectious diseases such as COVID-19 Evidence-Based Nursing 2021: 24:41
Nurses need accessible and thorough evidence-based guidance to support safe, effective care due to the risk of SARS-CoV-2/COVID-19 infection, sickness, and death. Selection and use of appropriate Personal Protective Equipment (PPE) is fundamental to respiratory protection programs and access to the latest evidence is vital to underpin practice and policy decisions.
The Evidence-Based Nursing commentary by Dr Dos Santos on a recent Cochrane Review of PPE use in healthcare workers offers some useful insights. However, there are other key details and findings from the original review that are also important for readers. While covering more of the body appeared to result in enhanced protection, more difficult donning, doffing, and poorer user comfort could lead to higher contamination risk. Also, more breathable PPE might result in a similar number of contamination spots on the user’s trunk compared to less permeable materials (MD 1.60, 95% CI −0.15 to 3.35) and potentially increase user satisfaction (MD −0.46, 95% CI −0.84 to −0.08).
Centers for Disease Control and Prevention guidance resulted in less contamination compared with using no guidance (small patches: MD −5.44, 95% CI −7.43 to −3.45). Other key findings include:
- One-step glove plus gown removal versus separate removal resulted in less bacterial contamination (RR 0.20, 95% CI 0.05 to 0.77) but not fluorescent contamination (RR 0.98, 95% CI 0.75 to 1.28).
- Double-gloving resulted in less viral and bacterial contamination (RR 0.34, 95% CI 0.17 to 0.66) but not fluorescent contamination (RR 0.98, 95% CI 0.75 to 1.28).
- Additional spoken instructions resulted in fewer doffing errors (MD −0.9, 95% CI −1.4 to −0.4) and fewer contamination spots (MD −5, 95% CI −8.08 to −1.92).
- Extra sanitation of gloves with quaternary ammonium or bleach decreased contamination, but not alcohol‐based hand rub.
- Additional computer simulation reduced doffing errors (MD −1.2, 95% CI −1.6 to −0.7).
- Improved skill scores following PPE donning video lecture (MD 30.70, 95% CI 20.14 to 41.26).
- Face-to-face instruction reduced non-compliance with doffing guidance more than folders or videos alone (odds ratio 0.45, 95% CI 0.21 to 0.98).
References
De Castella T. WHO says ‘at least’ 115,000 health workers have now died from Covid-19. Nursing Times. 27 May 2021. Available: https://www.nursingtimes.net/news/coronavirus/who-says-at-least-115000-h... (Accessed 31 May 2021).
Verbeek JH, Rajamaki B, Ijaz S, et al. Personal protective equipment for preventing highly infectious diseases due to exposure to contaminated body fluids in healthcare staff. Cochrane Database Syst Rev. 2020;4: CD011621. doi:10.1002/14651858.CD011621.pub4.
As fourth-year King's College London medical students, we have read this article with great enthusiasm and felt we could identify with the concerns raised. We want to offer an additional perspective.
In terms of direct hazards, we have noticed that some of our colleagues have felt unsafe on specific placements. This is undoubtedly the case in psychiatry where patients may, unfortunately, become verbally abusive and at times physically intimidating.
Additional concerns may arise when students are placed peripherally in areas where there is a substantial crime. This can involve anything from petty theft to more serious crimes like muggings.
A clinical risk that we have been well trained to anticipate and handle are needlestick injuries. This is an example of adequate preparation making us feel more comfortable, i.e. working with sharps despite risks involved.
There are times when senior staff expectations are exceedingly high, and due to indirect risks such as stress and burnout, us students may fail to reach these standards. In some instances, lack of support and understanding perpetuate the cycle of stress. We have, however, noticed that this is less likely to occur when students have a longitudinal supervisor.
Overall we understand that certain risks are unavoidable, and we appreciate the vast number of measures in place aimed at safeguarding against risks and hazards during clinical placements. From our experience as medical stu...
As fourth-year King's College London medical students, we have read this article with great enthusiasm and felt we could identify with the concerns raised. We want to offer an additional perspective.
In terms of direct hazards, we have noticed that some of our colleagues have felt unsafe on specific placements. This is undoubtedly the case in psychiatry where patients may, unfortunately, become verbally abusive and at times physically intimidating.
Additional concerns may arise when students are placed peripherally in areas where there is a substantial crime. This can involve anything from petty theft to more serious crimes like muggings.
A clinical risk that we have been well trained to anticipate and handle are needlestick injuries. This is an example of adequate preparation making us feel more comfortable, i.e. working with sharps despite risks involved.
There are times when senior staff expectations are exceedingly high, and due to indirect risks such as stress and burnout, us students may fail to reach these standards. In some instances, lack of support and understanding perpetuate the cycle of stress. We have, however, noticed that this is less likely to occur when students have a longitudinal supervisor.
Overall we understand that certain risks are unavoidable, and we appreciate the vast number of measures in place aimed at safeguarding against risks and hazards during clinical placements. From our experience as medical students, we find that these interventions are mainly aimed at direct risks and hazards with the indirect types being overlooked.
We propose that students be made more aware of the types of indirect hazards they may encounter and better equip them to escalate concerns. This can take many forms, such as lectures, debriefing sessions, reflective practice and of course, one on one meetings with longitudinal supervisors.
Am I muddled? Figure 3 is labelled as an example of 'high specificity'. Just following the simple rules set out for figure 2 shows that: specificity = number of true negatives / (true negatives + false positives) = 8/ (8 + 62) = 0.1111 or 11% & sensitivity = number of true positives (true positives + false negatives) = 30 (30 + 0) = 1 or 100%. Surely this is a highly sensitive test??
Comment from the Editor: Thanks to Dr. McDermott for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1 We hope it addresses all your concerns.
I may have missed something but the Figure 3 does not illustrate high specificity, it shows perfect sensitivity and very low specificity. There is the same confusion in the paragraph that explains it "There is a risk that a test with high specificity will capture some people who do not have Disease D (figure 3). The screening test in figure 2 will capture all those who have the disease but also many who do not. " That would be a test with high sensitivity. And the reference to the figures is confusing (I believe the authors are referring to Figure 3 all along).
"Comment from the Editor: Thanks to Dr. Lassale for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1"
I read this article with interest as I revised for an recent assessment. I have some comments which may be of use to you and your readers, as I feel there are mistakes in the use of the terms sensitivity and specificity, which seem to be interchanged at various points throughout the article, and notably in the second paragraph, where there are initially defined. From my understanding, the ability of a test to correctly identify a disease is sensitivity, and the ability to correctly identify the absence of a disease is specificity, rather than he reverse as presented here. This may be typographical, as the authors go on to correctly illustrate both sensitivity and specificity in the subsequent example, and in Box 1, however further instances of this error appear later.
Another prominent example is Fig 3 and the associated paragraph, which I believe refers to high sensitivity, rather than specificity. As specificity is TN/(TN+FP), a higher false positive rate would decrease specificity. The figure shows a test which is highly sensitive ((TP/(TP+FN) = 30/(30+0) = 1 = 100%), but with low specificity ((TN/(TN+FP) = 8/(8+62) = 8/70 = 0.11 = 11%).
There seems to be a further confusion, particularly in the paragraph using the Ottawa ankle rule as an illustrative example. The 'wide range of sensitivity' discussed early in this paragraph, is not sensitivity, but specificity as stated in the discussion of the cited article (8). Indeed, this is indicated in t...
Show MoreI think these have been confused in the "definitions"
"Comment from the Editor: Thanks to Dr. Nicholl for these comments, please see the corrected version of the paper at https://ebn.bmj.com/content/25/2/e1"
I am writing to highlight what I believe to be several fundamental errors in this article in trying to explain sensitivity and specificity, titled “What are sensitivity and specificity?”
http://dx.doi.org/10.1136/ebnurs-2019-103225
1) Partway through, the article states, "there is a risk that a test with high specificity will capture some people who do not have Disease D (figure 3)."
I believe this should state "high sensitivity", not high specificity, as this would be the complete opposite with low false positives. The same applies for the "high specificity" label for Figure 3 - this illustrates a high sensitivity test (with very low specificity as it gives many false positives), not a high specificity test.
2) In reference to the Ottawa ankle rules:
"They have been shown (in a systematic review) to correctly identify approximately 96% of people who have a fracture and to correctly rule out between 10% and 70% of those who do not have a fracture.(8) The wide range of sensitivity is likely to be due to differences in the education of the clinicians involved".
Assuming a positive test here identifies ankle fractures, I believe this is a wide range in 'specificity', not a "wide range of sensitivity" (which appears to be c.96%).
Sensitivity and specificity are repeatedly swapped incorrectly in this sect...
Show MoreThis reply is in response to the above published article on March 22, 2022. The article was very informative and captivating due to the focus being on relations to "The impact COVID-19 on nursing students". After reading this article I was able to understand the prevalence of mental health problems and sleep disturbance among nursing students during the pandemic. The pandemic has brought so much hurt to persons between losing a family or losing a job or anything beneficial to help them provide. The effect the pandemic has on students, has been very disturbing. As the article spoke about how transitioning has affected the mental health and sleeping problems, it was so relatable because in The Bahamas, student nurses and registered nurses have been affected drastically.
Show MoreThis article has given me a more global view on how COVID-19 made a negative impact. In the Bahamas, COVID- 19 devastated all forms of jobs and the people that make up the jobs. The results from the survey from the students inside this article showed how 80% were worried about the impact of transition and 62% were worried about management of the workload. In The Bahamas, nursing students were affected in several ways, such as not being able to find resources to do the classes, not being able to complete any labs or clinicals which became a set back and placed myself and others in a very depressing state. In the article under “As a nurse” due to the need of nurses, students were pushed at the f...
Response to: Dos Santos WM Use of personal protective equipment reduces the risk of contamination by highly infectious diseases such as COVID-19 Evidence-Based Nursing 2021: 24:41
Nurses need accessible and thorough evidence-based guidance to support safe, effective care due to the risk of SARS-CoV-2/COVID-19 infection, sickness, and death. Selection and use of appropriate Personal Protective Equipment (PPE) is fundamental to respiratory protection programs and access to the latest evidence is vital to underpin practice and policy decisions.
The Evidence-Based Nursing commentary by Dr Dos Santos on a recent Cochrane Review of PPE use in healthcare workers offers some useful insights. However, there are other key details and findings from the original review that are also important for readers. While covering more of the body appeared to result in enhanced protection, more difficult donning, doffing, and poorer user comfort could lead to higher contamination risk. Also, more breathable PPE might result in a similar number of contamination spots on the user’s trunk compared to less permeable materials (MD 1.60, 95% CI −0.15 to 3.35) and potentially increase user satisfaction (MD −0.46, 95% CI −0.84 to −0.08).
Centers for Disease Control and Prevention guidance resulted in less contamination compared with using no guidance (small patches: MD −5.44, 95% CI −7.43 to −3.45). Other key findings include:
- One-step glove plus gown removal versus separa...
Show MoreAs fourth-year King's College London medical students, we have read this article with great enthusiasm and felt we could identify with the concerns raised. We want to offer an additional perspective.
In terms of direct hazards, we have noticed that some of our colleagues have felt unsafe on specific placements. This is undoubtedly the case in psychiatry where patients may, unfortunately, become verbally abusive and at times physically intimidating.
Additional concerns may arise when students are placed peripherally in areas where there is a substantial crime. This can involve anything from petty theft to more serious crimes like muggings.
A clinical risk that we have been well trained to anticipate and handle are needlestick injuries. This is an example of adequate preparation making us feel more comfortable, i.e. working with sharps despite risks involved.
There are times when senior staff expectations are exceedingly high, and due to indirect risks such as stress and burnout, us students may fail to reach these standards. In some instances, lack of support and understanding perpetuate the cycle of stress. We have, however, noticed that this is less likely to occur when students have a longitudinal supervisor.
Overall we understand that certain risks are unavoidable, and we appreciate the vast number of measures in place aimed at safeguarding against risks and hazards during clinical placements. From our experience as medical stu...
Show More