Science topic

# Clinical Epidemiology - Science topic

Explore the latest questions and answers in Clinical Epidemiology, and find Clinical Epidemiology experts.
Questions related to Clinical Epidemiology
• asked a question related to Clinical Epidemiology
Question
Hello everyone!
We are developing a phase I randomized clinical trial, in 18 healthy volunteers, aimed to test the safety and pharmacokinetics of i.v drug. However, we want to test two different doses of the drug (doses A and B), and each dose is to be administered with a specific infusion rate: dose A will be administered at X ml/min, and dose B at Y ml/min.
We need to randomize the 18 patients with a 2:1 ratio (active drug vs placebo), in blocks of size 6. However, to maintain the blind, we also would need two different infusion rates for the placebo (X and Y).
What do you think is the best way to randomize the volunteers in this study?
One way could be to randomize the patients in a 2 x 2 factorial design: one axis to assign the drug vs placebo, and the other axis to assign the drug dose with the infusion rate. To maintain a 2:1 ratio for the first axis, a and 1:1 ratio for the second axis, in blocks of size 6. A second way could be to randomize "three treatments" (dose A with X infusion rate, dose B with Y infusion rate, and placebo), 1:1:1 ratio, in blocks of size 6, and then, to randomize patients assigned to placebo in blocks of size two (or without blocks) to infusion rate X or Y.
What do you think is the best manner to randomize in methodological terms? In the case of the first way, Do we need to test the interaction between dose and infusion rate? Do you have another idea to randomize the patients in this study?
Thank you so much for your suggestions and help.
Hi,
As there are only 18 units and a single dose study without repetition of treatments, Whynot go for simple randomization procedure with random number generator into the three groups (A,B,C) one of the groups is similar to other that gives a 2:1 ratio. Groups A,B and C could themselves be randomized.
Here is a good book on this subject.
Lachin, John M.; Rosenberger, William F (Wiley series in probability and statistics) Randomization in clinical trials: theory and practice [2 ed.] John Wiley & Sons,2016.
• asked a question related to Clinical Epidemiology
Question
Dear researchers,
I am analysing the relation between productivity and quality in hospitals, using performance indicators. The number of hospitals is not big, below 45 per year. Is it possible to broad research on multiple years using same hospitals more than one time? Certiainly, that will harm assumtion on independent observations. However, I am sure that that there is no (systemati or planning) intervention in order to change hospital performances.
What do you think about my approach?
• asked a question related to Clinical Epidemiology
Question
Hello everybody
As you know, due to the high cost and time-consuming of laboratory assays, like PCR, for the suspected cases of COVID-19, doing this type of investigation is not possible for all the referrals to medical centers with COVID-like symptoms. Even though in middle and low-income countries this issue will be more critical. As a result, the infection may be confirmed by the practitioners based on the manifested signs and symptoms, without requesting a laboratory assay, i.e. clinically confirmation of COVID-19.
Accordingly, my question is that if we have the COVID-19 data in both forms of "PCR confirmed and Clinically confirmed patients" in our dataset:
1.Should we ignore or analyze those who didn't confirm by PCR test?
2.Can doing so cause bias in the reporting results?
3.How should these two types be reported together in a paper? Merged or separated?!
I dont think so you rely on symptoms of covid instead of assay. This can obviously creating bias coz the symotoms of covid also overlaps with other diseases as well
• asked a question related to Clinical Epidemiology
Question
The general public must be made aware of the mode of transmission, presenting symptoms and the measures that can be undertaken to prevent the spread of infection.
Few options- Media, Webinars...
Awareness can be increased if everyone is involved, the task of making people aware of COVID-19 prevention should not be left to the government alone. Let all stake holders be involved: that is civic leaders, the church, the education bodies, traditional leaders NGOs and the others.
• asked a question related to Clinical Epidemiology
Question
I have a study that found an association between exposure to tricyclic antidepressants and the risk of preeclampsia. The number of women who were exposed and had an outcome (i.e. preeclampsia) was small: 210 exposed women and 10 of them developed late-onset preeclampsia. Generalized Linear Models with binomial distribution and log function was used to calculate the relative risk (SPSS software). The reviewer asked us to "report the model goodness of fit criteria (to ensure correct specification of the model)".
How should I reply him? Because our study is an exploratory study that suggested an association. We are not building any model or predicting anything. Besides, the number of exposed cases were too small to predict anything. Thank you so much.
Thank you very much for the advice. I am a pharmacist and am not very good at statistics. Therefore I used SPSS to handle. Please show me steps/guide materials to do it? Thank you :)
• asked a question related to Clinical Epidemiology
Question
Dear all,
I want to analayze the inattentiveness level in ADHD patients and compare the current score with the score of 6 months before now. I do not have any score of inattentiveness level of 6 months before now. If I ask the patient or the patients close relative to state or fill out a questionnare about inattentiveness level of 6 months before now( stating their behaviour of 6 months earlier), will the results be reliable? Will the comparison stay out of bias?
Patients' reports on Past Week is already questionable, let alone 6 months...especially when they have ADHD, a condition associated with commonly poor memory. I think you can safely assume that subjective reliability will be very low. If you don't have data from 6 months ago, ask patients the 6-month question anyway and in your report clearly state that the recall is only memory-based and provide reasons for potentially low reliability. If they report overall improvement, esp if they have improved grades to show it (assuming we're talking about students), use that as a part of your analysis.
• asked a question related to Clinical Epidemiology
Question
what is period of infectivity of NIPAH virus encephalitis and is there any specisic treatment?
Dear Dr.Sandip Das,
Nipah virus disease is an emerging zoonosis of public health concern. The infection was was time recorded in pig workers in Malaysia.The incubation period is variable (4 to 18 days). The disease is also reported from India.I have described this viral diease in my book entitled " Zoonoses" published in 2007.
We have published one paper on Nipah virus.
Pal, M. and Abdo, J.2012. Nipah virus disease: A newly emerging viral zoonosis. International Journal of Livestock Research 2: 65-68.
With kind regards,
Prof.Dr.Mahendra Pal
• asked a question related to Clinical Epidemiology
Question
Infection with coronavirus SARS-CoV-2 has affected every aspect of our life including scientific research. It is evident that all aspects of human subject research are potentially affected by the situation since recruitment and inclusion of participants/patients may be disturbed, characteristics of participants in ongoing studies may have changed and overall study protocols may have been flawed.
I would like the scientific community to reflect on these aspects including (but not limited to) the following:
How has the COVID-19 pandemic affected your research?
Has there been a focus shift in general interest, financial or otherwise?
Would it be necessary for studies to report in publications if and how the study was affected by the situation?
Which aspects need to be considered in the different fields of human subject research including (but not limited to) medicine, biology, psychology, and sports sciences?
Which statistical aspects need to be considered and how can we solve potential problems?
How were researchers themselves affected and do we see impact on other sciences?
Dear Marie-Madeleine Bernard, I appreciate your contribution, also with respect to mentioning the importance of monitoring side effects. I also agree that political interest has already affect research on COVID-19, some of these forcing premature data presentation/interpretation with strong effects on the final results.
• asked a question related to Clinical Epidemiology
Question
Hello dear Reserachers, Professors,
I'm asking myself and you, for the effects of Covid-19 on publication, reviewing process, and reclassification of priorioties on certain topics; e.g: and with high probability, the first topic will be all reserches treated this virus and relatid subjects...)
How you see this new shift ?
Best Wishes
I think we have to use the time in writing and reviewing. Currently, many of scientsts stay at home and the laboratory load is reduced. This may be a good chance to finish the delayed office works
• asked a question related to Clinical Epidemiology
Question
I performed a multivariate logistic regression to estimate the role of some baseline variables (e.g. age, sex, etiology etc.) on a long-term outcome (good, bad). I only have one patient group and no patient underwent any treatment.
Can I calculate absolute risk reduction (ARR) or number needed to treat (NNT) in this case?
Hi,
No, it is not possible. You need at least one group exposed to any factor of interest and one group not exposed to calculate absolute risk reduction.
Best regards.
• asked a question related to Clinical Epidemiology
Question
How is coronavirus transmitted?
How dangerous is coronavirus?
What are the symptoms of the coronavirus?
How is coronavirus diagnosed?
How long does the coronavirus live?
• asked a question related to Clinical Epidemiology
Question
In this article, we have 2 groups, patients with clefts and patients without clefts. Our outcome is teeth with developmental defects "enamel defects". The authors report this as a "cross-sectional study".
In this article, we have 2 groups, patients with clefts and patients without clefts. Our outcome is teeth with developmental defects or dental anomalies. The authors report this as a "retrospective study" (case-control?)
In this article, we have 2 groups, patients with clefts and patients without clefts. Our outcome is teeth with dental anomalies or developmental defects. The authors report this as a "cohort study"
My question is why is the study design different, if in all cases we recruit patients and examine clinically / with radiography to assess the outcome.
In the past (and sometimes still) clinical researchers sometimes used the term "case control study" to mean comparing outcome in an exposed group to that in a "control" (ie un exposed group) ,cross sectionally. That is why in modern epidemiology (ie since about 1985) the term case-referent study is now preferred .
• asked a question related to Clinical Epidemiology
Question
Can someone help?
I would like to run a mediation analysis in STATA with a survival model using a category independent variable (3 levels), a continuous mediator (or) a category mediator (3 levels), and a binary dependent variable (Yes/No).
Which command should I use in STATA?
Thanks.
• asked a question related to Clinical Epidemiology
Question
What are the data needed and exclusion criteria for the researches studying the prevalence of infectious diseases or viral infection incidence in epidemiological researches?
How to design the study and what are the suggestions to get a good questionnaire form?
Regards;
• asked a question related to Clinical Epidemiology
Question
Could fall down be the most common etiology of mandibular fractures especially in countries with rapid architectural urbanization? are there published articles or researches which support or have this result?
It is true in a country like India where due to rapid urbanisation & using of high velocity vehicles ,without proper road safety majors RTA is the commonest cause of Maxillofacial injury followed by fall.This is my experience in atertiary care centre like AIIMS Bhubaneswar.
• asked a question related to Clinical Epidemiology
Question
In terms of food interaction and confounding, are associations with health outcomes likely to be direct, or could it be explained by other factors such as decreased consumption of other foods? How would design a study to address this question?
In my opinion, I think it should be mixed effect, and food patterns analysis can apply to exam this mixed effects. Yes, you do factor analysis first to determine the patterns and use the factor score as dependent variables in regression model/
• asked a question related to Clinical Epidemiology
Question
Hi all,
I'm comparing the effect of different therapeutic methods on the one year survival rate of different groups of patients. I calculated Hazard ratio (HR) using this formula:
HR= [Ln(proportion of patients event-free on research arm)] /  [Ln(proportion of patients event-free on control arm)]
Also, I calculated Standard error (SE) of HR in different studies based on availability of data such as; confidence interval (CI) , p-value and number of patients at risk.
Unfortunately, in some studies, none of this information is available. So I wonder if you could help me with an alternate method to calculate SE using HR.
Chapter 32 from Rothman and Greenland Modern Epidemiology 2nd ed 1998 (p 643-673) deals well with this including dealing with missing SEs or CIs as long as you have denominator sizes in the individual studies.
• asked a question related to Clinical Epidemiology
Question
What is the type of the study that study a diagnostic tool such as CT , US and assess the accuracy , sensitivity and specificity . Would it be Cross sectional or observational
Many Thanks
How about a diagnostic accuracy study?
• asked a question related to Clinical Epidemiology
Question
Our paper published in the Journal of Clinical Epidemiology clearly outlines why doing this is not just inaccurate but actually wrong because it introduces bias. This is regularly done and advocated by Cochrane [1]. I think this erroneous practice should now stop - see link to the paper [2]
[1] Higgins JPT, Altman DG, Gøtzsche PC, J€uni P, Moher D, Oxman AD, et al. The Cochrane collaboration’s tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928.
[2] Stone J, Gurunathan U, Glass K, Munn Z, Tugwell P, Doi SAR. Stratification by quality induced selection bias in a meta-analysis of clinical trials. J Clin Epidemiol. 2018 Nov 17. pii: S0895-4356(18)30744-3.
Thanks Godfrey. Even a sensitivity analysis by quality seems questionable based on our results in the paper. So the Cochrane recommendations do not seem to be valid and we should attempt to bias adjust using methods that include all studies.
Thanks for reading my book :-)
Regards
Suhail
• asked a question related to Clinical Epidemiology
Question
Is meta-epidemiology a part of meta-research?
I understand the following facts:
"Meta-epidemiology is not the same as meta-analysis.
Meta-research in not the same as meta-analysis."
Meta-research includes all kind of secondary research in which data are pooled from primary published/grey/white literature on the subject; it can be either quantitative or qualitative and meta-epidemiology is a sub-type of quantitative meta-research.
• asked a question related to Clinical Epidemiology
Question
Its true that a specific maximum number can not be given, but I would like to know the possible implications of increasing the number of confounding variables. What factors would limit the number of confounds to be included?
Need to consider the appropriate adjustment sets required - I recommend starting with a system such as Textor's DAGitty: http://www.dagitty.net/ - and the ability to obtain the necessary data on an appropriate number of subjects for all required adjustment sets.
• asked a question related to Clinical Epidemiology
Question
I have a longitudinal retrospective data set of human medical records. They feature CONDITION and DRUG. There is no way of saying why a drug was prescribed other than observing the conditions/diseases present at the time.
I would like to know whether taking drug X has an outcome on a particular disease. The outcome will be the duration between repeated visits to the doctors. I have used a recurrent cox regression to classify whether a particular drug (as a covariant) is associated with a change in risk to the disease outcome.
The predictor/independent could be time to a particular reoccurring disease record (remember, this is recurrent so a little bit like migraine so the patient sees the doctor often) and the dependent/outcome variable would be some measure of the disease outcome. If I take e.g., 1250,000 patient records, align them so that the index date is defined by the particular drug of interest, I could be able to get a before-after effect.
I would appreciate any links, papers, tutorials, on an approach similar to what I am trying to do.
Thanks
The study
J ohn Edmeads, Helen Findlay, Peter Tugwell et al.:
Impact of Migraine and Tension-Type Headache on Life-Style, Consulting Behaviour, and Medication Use: A Canadian Population Survey.
Can J Neurol Sci Volume 20, Issue 2 May 1993 , pp. 131-137.
hits Your question: "Alleviate migraine remedy doctor consultations?"
Why do You need so many records for a study of clinical impact according to drug studies?
• asked a question related to Clinical Epidemiology
Question
A recent “Controversy & Debate” series in the Journal of Clinical Epidemiology suggests that the results and conclusions of nutrition epidemiologic research are both "pseudo-scientific" and “meaningless” (links below). This conclusion was based on the fact that FFQs and other memory-based dietary assessment methods (M-BMs) produce data that are “physiologically implausible” and have non-quantifiable (i.e., non-falsifiable) measurement error.
For example, there are myriad factors that render it impossible to ascertain if reported foods and beverages match the respondent’s actual consumption. These include reactivity, lying, false memories, forgetting, mis-estimation, pseudo-quantification, and invalid nutrient databases. Additionally, the use of M-BMs is based on multiple logical fallacies.
Thus, how can nutrition epidemiologic data be valid?
Carole, if you peruse my latest work, you will see that the "best papers" (as you called them) failed to cite or address the contrary evidence I presented in my last ~10+ publications on that issue (links to my most recent/relevant papers are below). As I wrote, reactivity, lying, false memories, forgetting, mis-estimation, pseudo-quantification, and invalid nutrient databases render data from FFQs and 24HRs meaningless.
More importantly, the evidence I presented were empirical, conceptual, and theoretical refutations of FFQs and other M-BMs and not mere limitations. I am sure you will agree that a refutation is much more than a mere "limitation". Thus, one cannot take "precautions" to ameliorate the effects of non-quantifiable errors and physiologically implausible data.
Frank, thanks for the reference. As I presented in my work below, there are many, many perspectives on why epidemiologic work in general, and the results and conclusions of nutrition epidemiologic research specifically, are "meaningless".
• asked a question related to Clinical Epidemiology
Question
I am after some suggestions on what statistical analysis I can perform to show a before-and-after effect in a longitudinal electronic healthcare record (EHR). I have N number of EHRs, of varying sizes/time-spans. Each record has a history of recurrent disease records (for the one disease). To see whether a particular drug has had an effect on the disease outcome (duration before the next relapse), I have used time-gap recurrent cox regression.
However, I would now like to see whether the disease outcome (a series of remissions into relapses, good = long durations in between, bad = short durations in between) is immediately clear from the first prescription of a particular drug. In my head I imagine, taking all of the records (of vary time-span sizes -- very important to remember), and adjusting so every record overlaps when the drug of interest is first prescribed. Y axis is disease prevalence or risk, and x axis is time. From before the initial drug prescription event, disease prevalence/risk should be high, then after crossing the initial prescription time, disease prevalence/risk should drop. This would help demonstrate the efficacy of the drug.
Some points to remember: 1) Each medical record maybe unique in timespan. 2) The first prescription event of a particular drug will happen at different times across the record set. 3) Some records may have no medical events before the drug was prescribed (as all the diseases of interest feel after the drug prescription of interest). 4) The number of medical events either before or after the first prescription of the drug may be sparsely populated (making binning by time very difficult) or richly populated.
Is there a name for this kind of analysis? I am using R. Any suggestions are very welcome.
Look like the data description fits into an interrupted time-series analysis
• asked a question related to Clinical Epidemiology
Question
My exposure is a continuous variable that has been measured at 9 follow-up examinations in each participant.
My outcome is a also a continuous variable that has been measured only once: at the end of study.
I would like to test whether changes in my exposure over time is related to the outcome.
What are the available statistical tests to evaluate such relationship?
I have used mixed model previously to test the association between an exposure at time x and change in outcome over time. In this case, the outcome was measured multiple times over time while the exposure was measured only at the beginning of study. I wonder if mixed model would work for modeling the changes in exposure over time as well? or are there better statistical tests to answer my question?
Thank you
Dear Faye Anderson,
I am aware of the tests that you suggested, however, I was wondering if there are any specific statistical tests that would allow modeling the changes in exposure over time.
For example, I have used mixed model previously to test the association between an exposure at time x and change in outcome over time. In this case, the outcome was measured multiple times over time while the exposure was measured only at the beginning of study. I wonder if mixed model would work for modeling the changes in exposure over time as well?
Thank you
• asked a question related to Clinical Epidemiology
Question
Prognostic predictors in clinical epidemiology
Hi I would like to comment that Lung cancer is the most common cancer worldwide. In 2008, the number of incident cases was estimated to be around 1.6 million (13% of all incident cancers). Mortality is high with 1.4 million of deaths the same year (18% of all deaths from cancer) (www.globocan.iarc.fr). Overall, survival rate at 5 years is <20% but heterogeneity is important and the search for prognostic factors has led to the publication of an impressive number of papers. However, due to the design and often retrospective nature of prognostic factors studies, few of these factors can really be used in routine care to guide management and to determine prognosis. More recently, with the development of so-called targeted therapies, more and more attention has been paid to the identification of predictive factors that might be better tools to guide therapy.
Prognostic versus predictive factor: A prognostic factor is generally defined as a factor, measured before treatment, that has an impact on a patient′s outcome “independently” of received treatment or of the general class of treatment. Populations of patients used for prognostic factors identification may be very broad (from resected stage I patients to stage IV patients scheduled to receive chemotherapy) or more specific such as patients treated with radical radiotherapy or stage III patients. Outcome is most often defined as overall survival but other outcome measures may be used, such as progression free survival, response to the anti-tumoural treatment as well as disease-free survival rate or proportion of patients alive at a specific time point. A predictive factor is a factor expected to be able to identify patients who will benefit from a specific treatment. The hypothesis that treatment effect is only subject to random variation does not hold any more. If a predictive factor is validated for a particular treatment, it will obviously guide therapy.
Prognostic factors:
1. Non-small cell lung cancer: The staging of cancer is one of the most reproducible prognostic factors with the TNM classification based on tumor size (T), nodal (N) and metastatic (M) involvement. The TNM system was first described by Denoix (1946) and successive TNM classifications for malignant tumors has been published since 1968 by the Union for International Cancer Control (UICC). The stage is one of the most used factors to guide therapy. According to the clinical stage, the survival rates at 5 years range from 50% for stage Ia to 2% for stage IV while, if the pathological stage is used, a 73% rate of survival at 5 years is observed for stage Ia decreasing to 13% for stage IV (table 1). Stage is a powerful prognostic variable summarising the information included in the three separate factors: T, N and M. Of course, taken separately, these factors are prognostic factors: an increasing tumour size worsens prognosis and the lymph node involvement is per se a major prognostic characteristic which has also an impact on the possibility of surgical treatment (N3 involvement being generally a contraindication to surgery). Pleural dissemination is a negative prognostic feature and, from the 7th edition, a patient with pleural dissemination is now considered M1a. In metastatic patients, a single metastatic site is less detrimental than multiple metastases.
2. Classical host-related and tumour-related factors: The second most reproducible prognostic factor, also very useful to guide therapy is performance status measured on the Karnofsky scale or on the Eastern Cooperative Oncology Group (ECOG) scale although its value has mostly been demonstrated for non-resected patients. Therefore, some authors have argued that chemotherapy for stage IV patients should be limited to patients with ECOG performance status 0 or 1 however, other publications suggest that some patients with PS 2 may also benefit from treatment.
3. Other biomarkers: There are plenty of publications in the literature about biological markers not measured routinely in clinical practice. Most often, these factors are not reproducible and their prognostic independent value is not proven, with adjustment for well-known prognostic factors. We will cite only those that have been studied with meta-analyses or pooled analyses of selected trials, although published data generally do not allow the study of the independent value of the possible prognostic marker. The following features have been suggested to be associated with a more favourable prognosis: p53 normal status, no EGFR expression, low microvessel count, low VEGF expression, no overexpression of c-erbB-2 with an effect possibly restricted to non-squamous histology, Bcl-2 expression; low KI67 expression ; absence of KRAS mutation ; TTF-1 positivity ; high level of p16 expression, low or no ERCC1 expression (advanced NSCLC treated with platinum-based chemotherapy) ; low class III β-tubulin expression, in resected patients ; low survivin expression, in resected patients only ; and low lymphatic microvessel density, in surgically treated patients . Regarding the prognostic value of angiogenesis, microvessel count was confirmed as prognostic factor in a meta-analysis based on individual data, only if assessed by the Chalkley method.
Metabolic factors: Numerous studies have looked at the prognostic value of tumor metabolic activity as measured by [F]-fluoro-2-deoxy-d-glucose positron emission tomography. These studies have been meta-analysed and this review has shown that high metabolic activity is indeed an univariate prognostic factor (estimated hazard ratio of 2.08). The independent value remains to be proven and the conclusion holds mainly for limited tumours as few stage IV patients were included in the published studies.
Prognostic classifications:
1. Small cell lung cancer: Small cell lung cancer is a highly chemosensitive tumour but progression-free survival and overall survival remain extremely poor. Long-term survival is rare and cure rate is reached in <5% of the patients . For years, treatment of small cell lung cancer has been guided by the extension of the disease: limited disease (generally defined as a disease limited to the hemithorax of origin, the mediastinum and the supraclavicular lymph nodes which can be encompassed in a radiation field) versus extensive disease. Respective median survival times range within 15–20 and 8–13 months. Recently, within the IASLC Lung Cancer Staging Project, data concerning 12,620 small cell lung cancer cases were collected and complete clinical TNM staging was available for 3,430 cM0 patients as well as complete pathologic TNM staging for 343 cases.
Predictive factors:
Development of targeted therapies is evolving rapidly for non-small cell lung cancer. With the term “targeted therapies” we mean a treatment that is supposed to target a specific characteristic of the tumour. This specific target is expected to be a predictive factor. Most of the research carried out on predictive factors in lung cancer has been devoted to non-small cell lung cancer and we will restrict this review to non-small cell lung cancer.
EGFR and TKIs: Tyrosine-kinase inhibitors (TKI) targeting EGFR, such as gefitinib and erlotinib, have been first tested in randomised clinical trials without patient selection in addition to chemotherapy, in chemotherapy-naïve or untreated patients. They failed to show any benefit of the TKIs, although some clinical factors were suggested to be predictive of benefit: Asian, female sex, non-smoking status, non-squamous histology. The true predictive factor was identified later; the subgroup of patients who benefit in terms of progression-free survival from TKIs were those with somatic mutations in the EGFR gene (exons 19 and 21).
KRAS and TKIs: The KRAS pathway links the EGFR pathway to cell proliferation and survival and KRAS mutations have been suggested as a mediating resistance to EGFR mediators. A retrospective analysis of the BR.21 trial, as well as a meta-analysis, confirmed that presence of KRAS mutation is a negative predictive factor for benefit of TKIs in advanced non-small cell lung (HR of 1.97, 95% CI 1.16–3.33 for KRAS mutated tumours, HR of 0.79, 95% CI 0.59–1.05 for wild-type tumours; p-value for interaction 0.003).
EML4-ALK and crizotinib: The fusion between echinoderm microtubule-associated protein-like 4 (EML4) and anaplasic lymphoma kinase (ALK) has been recently identified in a subset of non-small cell lung cancers. EML4-ALK is most often found in never-smoking patients with lung cancer. Its expression is mutually exclusive from expression of KRAS and EGFR; it has no prognostic value but it is a predictive factor for efficacy of the ALK inhibitor crizotinib. Early trials with crizotinib led to approval of crizotinib but confirmatory trials are still ongoing.
Predictive factors for chemotherapy activity:
Although chemotherapy drugs have not been developed with the hypothesis of the existence of a molecular characteristic to target, some studies have also searched to identify predictive factors that might be useful in the choice of a chemotherapy regimen. These studies are extremely important as chemotherapy remains a cornerstone in the treatment of early or advanced non-small cell lung cancer.
1. ERCC1 and p27 and adjuvant cisplatin-based chemotherapy in completely resected patients :
Adjuvant chemotherapy provides a demonstrated benefit in overall survival when given to resected patients but brings also some toxicities. It was hypothesised that not all patients benefit from adjuvant chemotherapy and some biomarkers have been studied in order to identify subgroups of sensitive patients. Among them, ERCC1 has been tested and it is suggested that patients with low or no ERCC1 expression do benefit from chemotherapy (HR 0.65, 95% CI 0.50–0.86) while those with high ERCC1 expression do not benefit at all (HR 1.14, 95% CI 0.84–1.55) with a significant interaction test showing that chemotherapy effect is indeed not the same across the two subgroups.
2. RRM1 in more advanced non-small cell lung cancer:: The predictive role of RRM1 for sensitivity to gemcitabine, an antimetabolite frequently used in combination with platinum has been recently studied in the context of a randomised trial comparing cisplatin, docetaxel and gemcitabine to cisplatin–vinorelbine. Although the analysis was retrospectively done on a subgroup of 261 patients (out of the 443 randomised), the results suggest, surprisingly, that the predictive role of RRM1 is present for sensitivity to cisplatin–vinorelbine with better outcomes observed for RRM1-negative patients (better disease control rate, better progression free survival (6.9 months versus 3.9 months; p<0.001), better overall survival (11.6 months versus 7.4 months; p = 0.002).
3. Gene signatures:: The signature proposed by Zhu et al. as prognostic might also be predictive of a benefit reached with adjuvant chemotherapy (cisplatin and vinorelbine) in stage IB and II resected patients.
Prognostic factors are very useful to get information about disease evolution and to construct homogeneous groups of patients. They can sometimes guide the therapy and identify subgroups of patients where more aggressive therapy is needed. They can also be used as stratification factors. They are however not powerful enough to be used at the individual level. Predictive factors are more directly useful in clinical practice as they are directly related to the efficacy of a specific treatment. A few of them now have a definite place for guiding therapeutic decisions in non-small cell lung cancer and we are on the way to a personalised medicine for the treatment of this disease. However, their development and validation are more difficult and may require very large sample sizes in particular when the incidence of the predictive biomarker is low. I am wishing that above Answer would explain your Question with Depth and Research. Regards
• asked a question related to Clinical Epidemiology
Question
I have 1600 drugs to treat a condition. I firstly test them each by performing a drug screen. I divide the 1600 drugs into successes and failures.
For example:
100 drugs were successful
1500 drugs were failures (had no effect).
I then have an in silico predicting model, to which I apply all 1600 drugs. Again, like the experimental drug screens, the predicting model will yield successful drugs and failure drugs.
For example:
X drugs were successful
1600-X drugs were failures
I want to know what kind of a statistical test I can perform that will return some measure of significance between the two methods used. So I can say whether both reproduce very similar success/failure outcomes (sets of drugs)?
Thanks
If I understand the problem correctly, you have two methods of evaluating the same "subjects" (i.e., drugs). Each method assigns an outcome ("works" or "does not work") to each drug. Thus the results of the two methods appear to be dependent, and so you should use McNemar's test for equality of proportions.
• asked a question related to Clinical Epidemiology
Question
Hi.
I'm interested to get in contact with researchers in Montana that has an interest in epidemiological research in general, or pulmonary/cardiovascular research with an epidemiological approach. Have tried online searches/googling, but no success that way, so why not try here?
Thanks Ahsan, I am specially interested in Amyotrophic Lateral Sclerosis, Multiple System Atrophy and Progressive Supranuclear Palsy. Also ME/CFS. I have mapped in great detail 100s of residences in relation to environmental sources of EMFs (MHz and GHz). Very high levels of Reactive Oxygen Species are generated as the body is in fact an electro-chemical construct as acts as a pick-up antenna for RF signals. We are about to have our third paper published in neurological literature.
Best wishes Anne Silk
• asked a question related to Clinical Epidemiology
Question
I have a medical longitudinal retrospective dataset, records between the observation period of 2000 and end 2016. For many reasons not every medical record spans that entire time-frame, e.g. the patient may have died, or they may have transferred in to the study half way through or transferred out at some stage.
A particular event (or exposure) is seen as a clinical event e.g., going to the doctor and saying or being told that you have a particular disease, e.g., a chest infection. That patient will also have a categorical variable to indicate whether they are a smoker or not.
I wish to count the frequency of chest infections per patient and distribute them over whether they smoke or not. I can imagine this would be a box plot with UQ and LQ being defined, frequency of disease on the Y, and a Smoke YES and NO on the X. This would be very easy to do. The problem I have though is that I am not sure how I deal with medical records of varying length. Surely there is bias if a smoker vs. non-smoker both have twenty chest infections, but there is a four year medical record difference?
Thanks
You are about to discover the concept of incidence density! This is the rate of events per unit time. In your case, the rate of events per 100 or 1000 person-years.
You can tackle the problem using Poisson regression, with length of observation set as the exposure time variable.
Alternatively, you can treat the data as time-to-event data with repeated events. This has the advantage that the probability of chest infections probably rises with age. You can use age as the time variable in the analysis, with subjects entering at the age they were first seen and exiting at the age of last follow up. In order to avoid immortal time bias, you need to declare them to be at risk from some point, say from age 18.
This is pretty straightforward in Stata, if you have it.
• asked a question related to Clinical Epidemiology
Question
I understand how survival methods can be used to determine the probability of survival given a dataset of 'time-to-events' with almost all examples considering cases of Alive/Dead e.g., cancer. However, how can I factor in cases of multiple remission and relapse events per person in a disease that will not take a life?
For example. Remission is defined by the absence for more than 90 days from medication or disease in a patient record. A relapse is returning to a similar medical/drug state any time after a single day beyond the 90-day disease/remit cut-off. To be considered as having ongoing treatment, there will be a continuous record of either drug prescription or disease code for less than 90 days at intervals (more than 90 days and we assume that the patient is in remission). e.g., visiting a doctor (or repeat prescription using the NHS model) at least once every three months.
Using these definitions, I can take an individual's medical history and table the number of days until time-of-event of diseased and remission. A good drug will mean remission was longer than diseased, or at least diseased is kept as short as possible even if there is then only a very short remission time. For example, Bob gets disease X at t=0 (and a drug prescription) and I start counting the number of days until there has been a 90-day absence of either, at which point I start counting the number of days as remission until the same drug or same disease appears and then I start counting again but for a diseased state.
patid days event
1 200 D (diseased)
1 450 R (remit)
1 340 D
1 500 R
2 ... D
2 ... R
I am using R and providing this data into the Cox regression function as though patid 1 (the first patient) is actually 4 people! Similar to how 4 people would be alive/dead in a cancer model.
I have coded all the logic to break down a group of individual's records into stages of diseased or remission. However, is it correct in a cox model (in R) to provide this information as is?
Unless I am missing something (which is totally possible), observations should be independent which would be difficult when using a single patient as though they are four different patients. So you will most likely need to change the way you have entered the information in to the cox model.
• asked a question related to Clinical Epidemiology
Question
i have developed a bioinformatic tool (Link: geltowgs.uofk.edu). that compares PFGE gel image analysis results to mathematical models of band sizes derived from WGS (FASTA files). I have suggested a new algorithm to count DNA fragments that co-migrates across each lane. We are about to publish ower work (Research gate DOI: 10.13140/RG.2.2.32752.76806) . The attached file illustrate our method.
thank you
sorry, the uploaded file was an old draft. The attched document is the complete one
• asked a question related to Clinical Epidemiology
Question
Currently trying to determine the feasibility of a three arm parallel group RCT pre-/posttest design on VAS pain intensity in osteoarthritis patients. Two intervention arms arms of different dosages of cannabis and an active control of codeine with analysis via ANOVA. However, there are currently no studies utilizing smoked/vaporized cannabis for the treatment of osteoarthritis or nociceptive pain so I don't know how I should calculate a cohen's d for the determination of a sample size. Any input or suggestions on finding an appropriate sample size or calculations would be greatly appreciated.
Cohen's d does not measure the size of the difference you expect to observe between the two treatment groups under comparison, but the minimum size of the effect you would like to observe in order to consider the difference as clinically relevant. You state you will analyse data by ANOVA, with one between-subject factor (treatment, three levels) and one within-subject factor (time, two levels), I guess. I imagine you are interested in evaluating the difference between the two dosage groups and the control group with respect to the pre-post variation in VAS measure (using a multiple comparisons test within the interaction treatment-by-time). You may compute the sample size referring to a Student t test for independent groups, setting two-tailed alpha = 0.0167 (applying the Bonferroni’s correction for three comparisons to the experimentwise alpha 0.05), power = 0.90, and different values for Cohen’s d. As an example, the number of subjects per experimental group would be n = 110 (total N = 330) for d = 0.5 (i.e. difference between the pre-post variation means equal to 50% of the pre-post variation SD) and n = 29 (total N = 87) for d = 1.0 (i.e. difference between the pre-post variation means equal to the pre-post variation SD). If you have an idea of the expected pre-post variation in control patients (mean and SD), you can estimate the absolute difference in pre-post variation corresponding to the different values of Cohen’s d. You can use G*Power (free download at http://www.gpower.hhu.de/) to calculate the sample size for different combinations of alpha, power and effect size. Best regards
• asked a question related to Clinical Epidemiology
Question
I would like to belong to this project. Which is the first step?
Sincerely.
Full Professor Faculty of Medicine Universidad Peruana Cayetano Heredia.
Internal Medicine - Geriatrics
Master in Clinical Epidemiology.
Mobile: 51 999395806
The intent is a mutually beneficial exchange of information and knowledge - so sharing will bring benefits. While the focus is on SAGE and SAGE-related data, we actively encourage comparisons to other data sets and interactions/exchanges on that to better understand the drivers of differences in health created by unique policy and cultural issues.
• asked a question related to Clinical Epidemiology
Question
Just for an undergrad research proposal, if you are doing an RCT looking at if an intervention (exercise) is effective at preventing a disease (hernia), is there any merit to continuing to follow-up subjects once they have acquired the disease if your outcome of interest is only incidence? It consumes more resources for follow-ups if it is not necessarily going to affect data analysis or anything..
It would seem to be an endpoint definition issue. If you define a fixed study period, and occurrence of any single hernia in the period constitutes failure of intervention, then I can't see any bias introduced by not following up after hernia occurrence, if you are nor interested in treatment or repeated /second herniae..
• asked a question related to Clinical Epidemiology
Question
Hi Everyone,
I am new to the latent class analysis technique and I'd like to ask if it would be possible to first run a latent class analysis on a sample of patients, followed by a agreement analysis between the LCA classification outcomes with the classification outcomes generated by an already established criteria (a priori)?
In other words, could a latent class analysis be used to verify the accuracy of established classification criteria for disease severity for example?
Another question, could we use the outcomes of latent class analysis to generate reproducible classification criteria that could be used in future studies?
Thank you
ASAIK, agree with the previous answer of Sandro. I would like to add only one thing: the classification criteria should be rationally interpretable, i.e. they should make sense not only as regards statistics but also from a clinical perspective. Otherwise, it might be difficult to use a classification in "real life".
• asked a question related to Clinical Epidemiology
Question
Hi,
I would like to compare the biochemical and cytological values like cell count, glucose levels, etc of test samples between two different tests, PCR and culture, for identification of Streptococcus pneumoniae. I would like to check if the values differed significantly among the samples positive by  each test alone, and in combination. What statistical test(s) should I use for this? Is Fischer exact test with two-tailed p < 0.05 can be used? Kindly help me.
Thank you.
Dear MURTHY
The choice of statistical analysis depends on the type of your variables ( scale, ordinal, ..) and the distribution of your data ( normal distributed data or not ), that is will determine your choice for parametric or non-parametric tests.
Good Luck
• asked a question related to Clinical Epidemiology
Question
I want to have a look at how the area-level effects of disease prevalence vary between ethnic groups.
Does it make sense to configure a random slope model to look at ethnic group (e.g. white or non-white)?
I did come across a similar question here: http://thread.gmane.org/gmane.comp.lang.r.lme4.devel/10096 The author suggests that in using a random slope on a dichotomous variable, it might aid interpretation to not use a random intercept: In my mind, in the context of my question, this would have the result of being able to see how the effect of ethnicity would affect disease prevalence, but no longer being able to see differences in disease prevalence between groups.
It is perfectly reasonable to use random effects when the predictor is a categorical one
Lots of examples in this
and this
and to be clear I am not talking about gender being the higher level unit but that the gender gap varies diffrentially over the higher-level units.
• asked a question related to Clinical Epidemiology
Question
I am currently working on a retrospective analysis of a patients database, which includes demographic variables, and other regarding healthcare (especially intra-operative and post-operative outcomes). However, it was not properly designed (6 years ago), therefore variables are not operationalized and some data is missing. Any recommendations to fix this old database and deal with the missing data?
Thank you.
Hello,
Are there any proxy measures you could use which might reflect the data for which you are searching?  For example, if you were looking for postoperative nausea and vomiting but some of the data were not coded for this, perhaps the administration of an anti-nausea agent implies the patient had this condition.
Best of luck on your study, the previous two answers are both quite valuable as well
Rich
• asked a question related to Clinical Epidemiology
Question
Hello all,
I am gathering information and opinions of the community about the most appropriate mouse strain to study development of wound infection. I am aware that it is highly important which bacteria/bacterium is studied. However, I am interested in information/opinions with respect of physiological responses to wound and infection. What are your experiences and pros/cons for particular strains (i.e C57Bl/6 vs BALB/c)
Looking forward to good discussions.
Cheers
The Murphy Roths Large (MRL/MpJ) mice provide unique insights into wound repair and regeneration. These mice and the closely related MRL/MpJ-Faslpr/J and Large strains heal wounds made in multiple tissues without production of a fibrotic scar. The precise mechanism of this remarkable ability still eludes researchers, but some data has been generated and insights are being revealed. For example, MRL cells reepithelialize over dermal wound sites faster than cells of other mouse strains. This allows a blastema to develop beneath the protective layer. The MRL mice also have an altered basal immune system and an altered immune response to injury. In addition, MRL mice have differences in their tissue resident progenitor cells and certain cell cycle regulatory proteins. The difficulty often lies in separating the causative differences from the corollary differences. Remarkably, not every tissue in these mice heals scarlessly, and the specific type of wound and priming affect regeneration ability as well. The MRL/MpJ, MRL/MpJ-Faslpr/J, and Large mouse strains are also being investigated for their autoimmune characteristic. Whether the two phenotypes of regeneration and autoimmunity are related remains an enigma.
• asked a question related to Clinical Epidemiology
Question
I'm planning to do a case-control study of Diabetes type 2 and lifestyle. It's gonna be a population-based study. do you have any suggestion ?
That is why in my earlier post, I had hinted upon the definition of defining 'Diabetes type II patients...
What would be the working definition...?
Siti's feasibility aspects really need to be given priority, but not at the cost of  unscientific selection of cases and control.
The criteria should be uniformly applied...
Look at 'resources' to finalize...
But definitely avoid plain symptomatic/ asymptomatic as a way to segregate cases and controls.
You may do some literature review of local or similar studies...
BUT,
Could you really specify your study objective...As that will clarify the appropriateness of such a case control  study selection.
• asked a question related to Clinical Epidemiology
Question
Proof of concept studies are usually small sample. I was wondering if there is any guidance regarding what power you should be aiming for.
Thank you for the resources! I know the general things about and I am not looking to conduct a study, I am re-analysing somebody else's triaks. So I am actually looking for a more sophisticated resource if possible applied to RCTs, where there are a lot of proof of concept trials done.
• asked a question related to Clinical Epidemiology
Question
About The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)statement
what is the possible score range of STROBE checklist? is there any cutoff score to define good or bad
Strobe statement is the checklist (recommendation of what must be reported and in what part of publication) and this is not an evaluation tool for publication quality assessment. If you need tool to assess quality of publications you can use GRADE method or the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies.
• asked a question related to Clinical Epidemiology
Question
Hello,
McHugh (2012) highlights some guidelines for cohen's kappa (quoted below), but I wanted to pose a question to others. What level of kappa do you usually see in published works? Or if you were a reviewer, what level of kappa would be your cut-off. I have some graduate students coding some data and the kappa statistics are ranging from .63 to .80. If you were reviewing, what would you think? Your thoughts and feedback on this would be appreciated. Thanks!
McHugh (2012)
"Cohen suggested the Kappa result be interpreted as follows: values ≤ 0 as indicating no agreement and 0.01–0.20 as none to slight, 0.21–0.40 as fair, 0.41– 0.60 as moderate, 0.61–0.80 as substantial, and 0.81–1.00 as almost perfect agreement."
Good afternoon,
if you were a reviewer, what level of kappa would be your cut-off. I have some graduate students coding some data and the kappa statistics are ranging from .63 to .80. If you were reviewing, what would you think?
First, are you having pairs of students rate your product?   Cohen's was designed for pairs of raters but he does not not state what level of agreement is 'acceptable'. (sees Cohen's original article (http://journals.sagepub.com/doi/pdf/10.1177/001316446002000104).
Hale and Fleiss looked at the distribution of the kappa itself (its variance).  It seems if there is a large variance with your calculated Kappa, the level of the Kappa might not be as meaningful. These authors describe the following, "moderate to high values of kappa, K = .50 and .75, under both sample sizes, 25 and 40"see http://www.jstor.org/stable/pdf/2532564.pdf
Here is a quote from Sim and Wright in Physical Therapy . Volume 85 . Number 3 . March 2005, page 264,  "The choice of such benchmarks,however, is inevitably arbitrary,' and the effects of prevalence and bias on kappa must be considered when judging its magnitude"
If you are using more than 2 rater's to evaluate a product, you might consider Fleiss' Kappa looks at multiple raters and he also provides a formula for the standard error of kappa.  This last point could help you determine the degree to which your ratings are good or poor.(see http://www.wpic.pitt.edu/research/biometrics/Publications/Biometrics%20Archives%20PDF/395-1971%20Fleiss0001.pdf
Have a great day!
Rich
• asked a question related to Clinical Epidemiology
Question
As you know we have clinical experience using NAVA, so we could participate in a multicenter study ;-)
Hello  Guillaume, in our PICU in Fernando Fonseca Hospital we also have experience with NIV-NAVA. Do you need another partner?
best regards
Helena Loureiro
• asked a question related to Clinical Epidemiology
Question
Why invasive infections due to C. tropicalis more common in India and tropics ?
Genus is Candida , we need to explore the details on consumption of azoles. Thank you geraldine
• asked a question related to Clinical Epidemiology
Question
I am performing a bibliometric analysis on the field of cancer.
I have found many studies which are derived from human samples and which would fall under the category of "molecular epidemiology" such as Genome-wide association studies -GWAS- or genetic risk prediction studies -GRIPS-, etc.
But how would they be best classified under the traditional concept of epidemiological design? Could they be classified as an "observational" or "interventional" study, or they need be considered as a different study design under the concept of "molecular epidemiological study"?
Thanks!
They would be classified as you would any other epidemiologic study, depending on the study design: observational, case-control, cohort, retrospective, prospective, etc.  The molecular element is only a tool to investigate something, such as genes or biomarkers.  Commonly these studies are case-control or observational or (often small)  followup cohort studies because they look at the molecular (or genetic) makeup of people and follow their outcomes or look (usually retrospectively) at their risk or odds of having a disease.
• asked a question related to Clinical Epidemiology
Question
I am working on a systematic review and meta-analysis of RCTs and I am wondering if there is a different approach for rating the quality of the evidence if the difference is statistically significant versus when the difference is not. Or should the quality-rating be restricted to only the significant outcomes? To me it does not seem logical to downgrade the quality of the evidence for an outcome (with a non-significant result) because there is a high risk of bias in the included studies.
Hi Bert,
1. Quality of evidence is to be rated for all outcomes of interest irrespective of their results - statistically significant or not. Quality rating should not be restricted to significant outcomes.
2. If the pooled results are not significant, the quality of evidence gets usually downgraded under the heading 'Imprecision'. For example, if the pooled RR (95% CI) is 0.6 (0.1 to 1.1), we downgrade the quality for being imprecise.
3. But this is not true for all non-significant results. For example, a pooled RR (95% CI) of 0.98 (0.92 to 1.08) will not be downgraded for 'Imprecision' because the result is  precise.
Downgrading the quality of evidence for risk of bias is a completely different thing. It has nothing to do with the significance of the pooled results.
I hope this clarifies your query.
Jeeva
• asked a question related to Clinical Epidemiology
Question
I am looking to sample questionnaire which measure knowledge, attitude and practice to predict risk behavior for cardiovascular diseases.
Hi,
I agree that the WHO STEPS modules would be simple tools for risk assessment
• asked a question related to Clinical Epidemiology
Question
Recently, we completed a TST survey among diabetic population in some Malaysian primary care clinics. We used 2 tuberculin unit (2 TU of RT23) during the TST survey. I would like to know how the use different tuberculin unit (2 TU, 5 TU, 10 TU) may affect the TST results and how the results should be interpreted. Are there any evidence in literature to substantiate 2 TU reduces false positives in a high burden country with wide BCG coverage?
In Spain it is used only the dose of 2UT and it does not seem to exist any basis to use other in the literature. In those cases where this dose has been used and previously vaccinated with BCG, there is a high probability of a false positive. Therefore, it is useful and it is what we do to determine IGRA.
I hope that this reply serves you.
Best regards.
• asked a question related to Clinical Epidemiology
Question
There are only a few studies on epidemiology of pediatric thyroid cancer mostly restricted to adolescents and less children (<15). Thyroid volume is smaller in children. Thus, the commonly used T-Staging in adults may underestimate the impact of tumor size on prognosis in children.
Dear Jamshid;
Thank you so much for putting forward this interesting question. You concern about underestimation of PTMC risk is quite real in children.
1- PTC microcarcinoma is rapidly increasing in the world mainly due to availability of high resolution ultrasonography and density of endocrinologists! Children is not away from this inadvertent screenings. Ito et al from Japan showed in prospective study of >1200 patients with PTMC that surgery should not be offered to >90% of these patients. Interestingly in that report, younger patients (<40 yrs) had more chance to have progression of disease compared to  others. this is concordant with the finding that the higher the age of the patients, the higher the risk of developing thyroid nodule and PTC. So we may conclude that PMTC has more chance of progression in children than older patients.
2- Multifocality, lymphnode metastases and distant metastases is nearly two times higher in children than older patients. This finding again emphasis on more scrutiny in children in PTMC.
3-Actually TNM classification is very limited role in pediatric patients, as all patients would be categorized as stage I or II (in case of distant metastasis). Further more TNM staging is designed for prediction of death which is extremely rare in this patients. So risk stratification aimed for prediction of risk of recurrence or persistent disease is much more useful in these patients. In my own experience, first Tg( Thyroglobulin measured 6 weeks after surgery and before radio-ioidne therapy) is the most reliable predictor of recurrence or therapy failure in these patients. So I do not rely on size and I prefer to look at the neck by ultrasonography and check first Tg  and Anti-Tg Ab in off T4 state in all pediatric patients with PTC before considering it as a real PTMC,
With best regards
• asked a question related to Clinical Epidemiology
Question
A text that explains the concepts without much math and formulae? Beginners in medical research in clinical trials/epidemiology often need a basic book on medical statistics that is appropriate for self study.
The Stanford university online course on medical statistics is very good (and free).
• asked a question related to Clinical Epidemiology
Question
The literature on wax and deafness is sparse, subjective, contradictory and confusing.  My clinical experience in a tertiary ENT clinic agrees with that of Politzer (1908), in that hearing loss only occurs when the meatus is completely obstructed by cerumen, or if a small passageway through is blocked off when water gets into the meatus. The literature confounds impacted wax with meatal occlusion. So is there any reliable data on the population attributable fraction of deafness due to wax, or data on hearing tests before and after removal?  (There are reasons other than possible deafness for removing wax).
My clinical experience: Even if you have a free passage in the ear canal with a small visualisation of the tympanic membrane + a normal tympanometry you will most likely have a conductive effect on 6-8 kHz. Take it for granted! With an occluded wax plug you cannot rely on the PTA if there is any form air-bone gap found on the audiogram. Always tympanometry to find out if it´s occluded. Removal of wax is crucial for PTA.
If the question is referring to sensorineural components in long term conductive loss, there are some articles on the subject but haven´t read them more than the abstract so cannot discern on this matter. One example:
Liberman, M. C., Liberman, L. D., & Maison, S. F. (2015). Chronic conductive hearing loss leads to cochlear degeneration: E0142341. PLoS One, 10(11) doi:10.1371/journal.pone.0142341
An animal study about cochlear degeneration from chronic conductive hearing loss. Not exacty the same subject but could probably gain some knowledge about the neurophysiological consequenses of conductive hearing loss.
• asked a question related to Clinical Epidemiology
Question
Dear all,
Can anyone share or help to find a good systematic review of cases of retained foreign bodies after previous C-sections? Thanks a lot!
• asked a question related to Clinical Epidemiology
Question
Im involved in a project which we desired to answer the follwoing: what is the probability of this patient to be sucessfully free of ventilation support 2 days from now, or 4 days from now or a week from now, if the ventilation support team starts its removal right now?
I have data from a cohort where many predictors, such as ventilator sets, blood gases etc, are measured every two days up to ventilation support successful removal. Therefore, the dataset has a typical predictors time dependent structure.
The first question is: is it feasible to use such data to develop a prediction model? I looked around and found some stuff supporting in the positive answer.
But I also found some stuff supporting a negative answer:
So far, a believe that it is possible, and in such a model the outcome prediction can be made at any time, not only at baseline as in traditional right censored survival model. Am I correct? However, Im not completely sure that cox proportional hazard model is adequate for this purpose. If it is not, what are the alternatives?
But then other questions follow.
To do time dependent predictions, I guess I must set the baseline date of every single patient to 1,  and the following dates where predictors are repeatedly measured must be set as differences to his/her baseline date, in a way that, similar to the traditional right censored model, all patients will start at day 1. Is this reasonable?
The last question. Suppose I have developed and validated this model, and I have five different risk groups as in this paper
If the patient is classified at group low risk, but after a while in this group without the outcome, he/she changes the predictor values and this moves he/she to another risk group. Does the risk estimation for her/him will be like just jumping from one survival curve to another at the same time? Or his/her time will be set to baseline time in the survival curve as he/she are just starting the follow up right now?
Another way to ask the same thing...the patient is at baseline and I want to estimate the probability of the outcome at day 2 from now. This is simple. Now suppose that he/she did not have the outcome and up to day 2 of the follow-up. Now I want to predict the outcome in the next two days with different predictors values. Should I predict for 2 days or should predict the outcome at day 4 of the follow-up?
I was challenged with this question. For me, so far, the correct way is to predict at day 4 (or just jump from one survival curve to another at the same time), as it correspond to "calendar" date of that patient in the follow-up, as it is in the dataset. However, I heard an argument that made me wonder. This argument is that survival models deal with conditional probabilities in a way that if this patient did not have the outcome up to now, the probability for the next two days are conditioned to his survival. Therefore, predictions should be made for his next two days, not at day 4.
Any light?
Dear beatrice,
Thank you for your feedback. I did read some stuff you posted, however, at the moment, my concern are the survival or other dynamic model suiting to answer my problem of individual predictions and not practice on ventilation support practice. The stuff I read, did not give me a different light so far. It turns that after a while banging around, I found some answers I was looking for in Therneau's book on survival analysis. Yes it is possible to make individual predictions with time dependent predictors. To set or not to set the baselines dates of each patient is a choice whether to consider a calendar date or to assume that there is no cohort effect and all patient start counting the time from a certain event such as hospital admission or aptitude of ventilatory support removal. I my case It would be more interesting set a baseline for everyone instead of using a calendar date. But there is some particular stuff to consider. At last, the individual predictions will always count from time zero of each patient. This means that, if the analysis considers risk groups, and the patient, at some time at the follow-up, has a change in the predictor value, he jumps from one risk group to another at that particular time. Other issues rise, for example, suppose I am at baseline and I want to make individual predictions at day 2, 4 and 7. I will have to guess reasonable values for the time dependent predictors at the days 2 and 4 (assuming I have the values from the baseline) . At first, carry over may be reasonable strategy, but updating the prediction will be required as soon as the information changes as it may substantially change the following survival probabilities. Another issue, is that Im not sure so far what calibration and performance measures are up to the validation phase of such a time dependent model, as, besides R2, all other measures that I found only works for the traditional right censored data. I will eventually read more of the stuff you sent to become more aware of others think and conduct analysis on this subject. Regards.
• asked a question related to Clinical Epidemiology
Question
R-Lipoic acid, acetyl-L-carnitine, N-acetyl cysteine plus other agents may be important in protecting mitochondria, but making sense of the literature is challenging: any light that might be spread would be greatly appreciate.
Dear Lewis,
I have searched the Cochrane Library and I found 7 systematic reviews of the literature and a protocol on the use of lipoic acid in several disorders (a list is attached below)
Best regards.
Record #1 of 8
ID: CD004244
AU: Klugman Anthony
AU: Sauer Justin
AU: Tabet Naji
AU: Howard Robert
TI: Alpha lipoic acid for dementia
SO: Cochrane Database of Systematic Reviews
YR: 2004
DOI: 10.1002/14651858.CD004244.pub2
Record #2 of 8
ID: CD002779
AU: Zakrzewska Joanna M
AU: Glenny Anne-Marie
TI: Interventions for the treatment of burning mouth syndrome
SO: Cochrane Database of Systematic Reviews
YR: 2005
Record #3 of 8
ID: CD004573
AU: Ang Cynthia D
AU: Alviar Maria Jenelyn M
AU: Dans Antonio L
AU: Bautista-Velez Gwyneth Giselle P
AU: Villaruz-Sulit Maria Vanessa C
AU: Tan Jennifer J
AU: Co Homer U
AU: Bautista Maria Rhida M
AU: Roxas Artemio A
TI: Vitamin B for treating peripheral neuropathy
SO: Cochrane Database of Systematic Reviews
YR: 2008
Record #4 of 8
ID: CD004426
AU: Pfeffer Gerald
AU: Majamaa Kari
AU: Turnbull Douglass M
AU: Thorburn David
AU: Chinnery Patrick F
TI: Treatment for mitochondrial disorders
SO: Cochrane Database of Systematic Reviews
YR: 2012
Record #5 of 8
ID: CD005492
AU: Mirza Nasir
AU: Cornblath David R
AU: Hasan Syed A
AU: Hussain Usman
TI: Alpha-lipoic acid for diabetic peripheral neuropathy
SO: Cochrane Database of Systematic Reviews
YR: 2015
Record #6 of 8
ID: CD010470
AU: Kumbargere Nagraj Sumanth
AU: Naresh Shetty
AU: Srinivas Kandula
AU: Renjith George P
AU: Shrestha Ashish
AU: Levenson David
AU: Ferraiolo Debra M
TI: Interventions for the management of taste disturbances
SO: Cochrane Database of Systematic Reviews
YR: 2014
Record #7 of 8
ID: CD008943
AU: Chaparro Luis Enrique
AU: Wiffen Philip J
AU: Moore R Andrew
AU: Gilron Ian
TI: Combination pharmacotherapy for the treatment of neuropathic pain in adults
SO: Cochrane Database of Systematic Reviews
YR: 2012
Record #8 of 8
ID: CD011546
AU: Livingstone Nuala
AU: Hanratty Jennifer
AU: McShane Rupert
AU: Macdonald Geraldine
TI: Pharmacological interventions for cognitive decline in people with Down syndrome
SO: Cochrane Database of Systematic Reviews
YR: 2015
• asked a question related to Clinical Epidemiology
Question
Dears researchers,I want to analyse association between disease(absence or presence :dependent variables ) and SNP (independents )and others parameters  by using logistic regression binary (spss),please how can i do adjustment for age and sex. Thanks
Your questions reveal you are new to statistical modeling including logistic regression. That’s all right. Every one has to start somewhere.
To make it simple, I will start explaining how you interpret the results with only SNP and sex as covariates.
You have told us that you have coded women as “1” and men as “0”, and you have probably coded the presence of SNP as “1” and the absence of SNP as “0”.
In you logistic regression you want to predict the odds ratio of disease (coded “1”) dependent on the presence of SNP, without and with control for sex:
Without control for sex your model is simple to include the SNP variable as the only explanatory variable. If the presence of SNP predict the disease you will find an odds ratio higher than 1, with an associated p-value less that 0.05 (or a 95% confidence interval not including 1).
Now you want to control for sex: You do that simply by adding the sex-variable to your model so that both SNP and sex are explanatory variables. BUT before you do that it would be a good idea first to see if you get what you expect when you include sex as the only explanatory variable. If women have higher risk of disease than men, then again you will find an odds ratio higher than 1, with an associated p-value less that 0.05 (or a 95% confidence interval not including 1).
Now you are ready to see what happens when you include both SNP and sex as explanatory variables. Let’s imagine you find an odds ratio for SNP = 7 and an odds ratio for women = 2.
Then the interpretation is this:
• The odds of disease for men without SNP is set to “1”. This is because men without SNP with the coding you have chosen automatically becomes the reference group (sex=0 and SNP=0).
• The odds for disease for women without SNP compared to men without SNP is “2” (that is, the odds ratio found for sex).
• The odds ratio for men with SNP compared with men without SNP is 7 (that is the odds ratio fond for SNP).
• The odds ratio for women with SNP compared with women without SNP is also 7 (that is the odds ratio fond for SNP).
• Finely the odds of disease for women with SNP is 14 (that is the odds ratio for sex multiplied by the odds ratio for SNP (2x7)).
The above interpretation is based on the assumption that there is no interaction between SNP and sex. That is, that you assume that the effect of SNP is multiplicative, and is the same for women and men. This might not be true. To check for interaction you will need to also include an interaction term between SNP and sex in your model. In SPSS there is probably a smart way to do that, but to make it quite clear, you can make such an interaction term yourself: You need to generate a new variable which is simple the product of the SNP and the sex variable. E.g. SNP_sex= SNP x sex. This new variable will take on the value “1” whenever a woman has SNP and “0” for everyone else.
Now you repeat the previous model but now you also include the new variable “SNP_sex” in the model.
If the odds ratio for this interaction term is significant, the interpretation of your data will be another. Let’s say you get the following results:
Odds ratio for SNP:                       8                           p=0.03
Odds ratio for sex:                         1.8                       p=0.02
Odds ratio SNP_sex:                    0.8                       p=0.04
Then the interpretation is this:
The effect of SNP for men is an increase in the odds by a factor of 8, whereas the effect of SNP for a women is only 8 x 0.8=6,4, and the odds of disease for a women with SNP as compared with a man without SNP is 8 x 1.8 x 0.8 = 11,52.
If the odds ratio for the interaction term is in-significant, the interpretation of your data will be that from the previous model without the interaction term included.
In the example above, the odds ration for the interaction term was less than 1 (OR<1) which means that the effect of sex and SNP is not fully multiplicative, that is, sub-multiplicative. You could also find that the effect is supra-multiplicative, that is OR>1.
Regards Kim
• asked a question related to Clinical Epidemiology
Question
Prompt please, what modern methods of analysis of time series used in morbidity statistics? There are scientific articles with examples on this subject? Thank you!
• asked a question related to Clinical Epidemiology
Question
I'm having trouble with a project. I am using ONS deprivation indices (https://www.gov.uk/government/statistics/english-indices-of-deprivation-2015)
which assign deprivation indices according to postcode/LSOA.
My problem is that I'm unsure what to do regarding homeless people. Since they have no postcode, I can't assign a deprivation score. Homeless people are a 'deprived' population, however the deprivation indices is an area level index. One option would be to exclude them from the deprivation analysis, but this would then underestimate the impact of 'deprivation'.
Would be interested to hear from anyone with experience in this.
I have tried categorising in both ways. Unfortunately with the dataset I am using (RCGP RSC) it is not clear whether postcode is missing because of incomplete data or because of homelessness which makes this more difficult. With most missing data adding an extra category seems to work well e.g. categorising smoking status as; current smoker, ex-smoker, smoker, and not recorded. This way no data is lost. As you would expect the outcomes of 'not recorded' smokers falls somewhere between smokers and non-smokers. If you are categorising IMD for your regression models then adding an extra category for homeless would be a straightforward approach.
• asked a question related to Clinical Epidemiology
Question
We are performing a systematic review regarding studies that validated administrative databases for ICD-9-CM or ICD-10 codes for different cancers. There could be events in which there are algorithms that were developed to identify multiple cancers.
Have a look at IARC Tools program and also SEER manual on multiple primary tumours. See:  http://seer.cancer.gov/tools/mphrules/
• asked a question related to Clinical Epidemiology
Question
I have been documenting strange step-like changes in deaths in a number of countries and would like others to check and see if these observations can be replicated using small-area death statistics. Attached is a paper documenting the parallel effects of these events on medical admissions to hospital and it gives an idea of the sort of analysis which could be required.
If needed many of the supporting studies can be accessed at www.hcaf.biz in the 'Emergency Admissions' web page - which also contains the published stuidies on deaths.
Much appreciated if you can assist.
Dear Arthur, I have demonstrated that the increase in both deaths and medical admissions associated with these infectious-like events is diagnosis specific. Hence while analysis of all-cause mortality picks up the timing of the events the obvious next step is to analyse trends for specific conditions/diagnoses. I have done this in a preliminary way in several of the papers listed in the Emergency Admissions folder at www.hcaf.biz which can be downloaded. Have a quick read. Cheers Rod
The attached book chapter may be helpful. Type in 299174 to open a printable version. I have purchased the right to distribute this chapter.
• asked a question related to Clinical Epidemiology
Question
We are more versed in mercury, manganese, and lead but have a large dataset (n = >4000) that also has data for cadmium. We found several significant demographic associations and also have further CBC, folate, other blood data that can be further analyzed
I would like to join the team, contact me.
• asked a question related to Clinical Epidemiology
Question
I plan to do a systematic review to summarize the evidence regarding the risk factors of a disease. I have 3 main questions.
1) Should I include longitudinal studies only or both cross-sectional and longitudinal studies that investigated the risk factors for a disease?
2) Can you recommend some risk of bias assessment tools for evaluating this type of systematic review?
3) Apart from odds ratio, relative risk and hazard ratio, are there any other statistical key words that I should use for my literature searches so that I can include all the relevant papers?
Thank you very much!
Arnold
It's helpful if you can do a scoping exercise to get a feel for what's out there. Usually you'd have to do this when working up the systematic review protocol. As already mentioned above cross-sectional studies can't indicate predictive risk, only an association. If there are sufficient RCTs then you might justify focusing only on RCTs but if not then it would make sense to include other longitudinal designs. If you do include multiple study designs then you should use different critical appraisal tools to cover these. Downs and Black and Newcastle-Ottawa as already mentioned above are useful but not without limitations. When selecting a critical appraisal tool it may help to pilot test it on a subset of studies (e.g. some identified in scoping) as this may identify whether it needs modifying for your particular question and can sometimes highlight issues not foreseen. Deeks and colleagues reviewed 194 different tools for critically appraising non-randomised studies, including those mentioned already (Health Technology Assessment 2007; 36: 666-676). Although a fairly old paper now I think this is still useful as a general guide due to the range of tools covered. New tools are continually being developed, e.g. the Cochrane Collaboration has developed "ACROBAT-NRSI" for non-randomised studies although it is a beta version at present and in my opinion is quite a "clunky" tool to use because it is fairly complex (I think some general principles underpinning that tool are really useful  though - but it is probably too early to use this approach since more testing and validation would be needed). Of course, there is considerable flexibility with any tool to modify it so long as you clearly report how and why this was done. It is important to justify why you used any particular tool(s) and to be honest about any limitations, noting this in the systematic review report.
Regarding the outcome measures to include in searches, you might find that you don't need to specify this level of detail in the search strategy - a pilot test of your draft strategy to see whether it picks up the outcomes you need can be very valuable. If you were to find large numbers of bibliographic records which don't report relevant statistical outcomes then perhaps you could try adding statistical terms in to the search strategy to limit the searches, but you would have to be confident that you wouldn't be excluding studies that don't report statistical outcomes in their title/abstract (or whatever fields you are searching).
Hope your review goes well... Good Luck!
• asked a question related to Clinical Epidemiology
Question
Or for that matter any oral disease...
Search as "dermatoglyphics and periodontal diseases" in Google scholar and you would get the articles and answers.
Regards
• asked a question related to Clinical Epidemiology
Question
I specifically need those which assess reliability and reproducibility studies. Thanks.
When writing systematic reviews or meta-analyses one needs to evaluate the quality of the studies before deciding which studies to include in the review. There are different criteria used for different types of studies: RCT, observational studies, genetic studies economic evaluation studies, studies about assessment instruments etc. There is a whole literature on 'reporting guidelines' for these different types of studies. A couple of links are as follows:
uniform Requirements for Manuscripts.  www.ICMJE.org
randomized controlled trials: CONSORT guidelines(http://www.consort-statement.org
observational studies (cohort, case-control or cross-sectional designs): STROBE (http://www.strobe-statement.org)
genetic association studies: STREGA guidelines (PLoS Med 6(2): e1000022. doi:10.1371/journal.pmed.1000022)
studies of diagnostic accuracy: STARD guidelines (http://www.consortstatement.-org/stardstatement.htm)
systematic reviews and meta-analyses: PRISMA guidelines (http://www.prisma-statement.org)
meta-analyses of observational studies in epidemiology: MOOSE guidelines
To assess the quality to studies about the theraputic interventions in medicine, the Cochrane Collaboration (which has developed the GRADE criteria) is the most authitative source.
• asked a question related to Clinical Epidemiology
Question
We are seeing lot of viral thrombocytopenic fevers which test negative for Dengue serology and even PCR but have a clinical picture similar to Dengue.
Because we have ruled out bacterial infections with appropriate tests and left with a huge chunk of patients with clinical picture similar to Dengue fever
• asked a question related to Clinical Epidemiology
Question
I am developing a tool to predict the age of onset of Alzheimer's disease using Bayesian statistics to estimate genotype. I have come up with a proportional hazards model using simulated data based on existing literature. I am using covariates APOE4 status, sex, history of TBI, history of DM and education level.
Potentially I need two datasets: The first would be the APOE4 genotype for the subject and familial history ie current age (or death) and age of onset for the subject, parents and grandparents.
The second dataset I would need is age of onset and genotype, sex, Hx of TBI, Hx of DM, and education level.
I would prefer an anonymous dataset(s). I would be happy to collaborate with someone on this project.
There is the Alzheimer's Disease Neuroimaging Initiative that might be of some help.
• asked a question related to Clinical Epidemiology
Question
I am currently looking into studies for epidemiological data of offshore workers outside the US in order to use it for a project looking into the nature of medical cases on offshore installations and their causes.
thanks everyone
• asked a question related to Clinical Epidemiology
Question
There is significant amount of hospital based research with loads of biomedical research institutes. I am curious to know if there is evidence showing the benefits of hospital based research to the health of populations, training of health workers and clinical care (the easy bit).
The advantage of hospital-based research is that the technical manpower & laboratory facilities are available to diagnose Illness and do research at hospitals . Regarding Infectious diseases , symptomatic ( mild & severe ) patients can be evaluated & treated in hospitals only . The Incidence data of any Infectious disease can be easily obtained from hospitals . The prevalence data can be designed for community , by utilizing the hospital data . Most of research publications on severe Infectious data come from hospitals only .
My personal experience on human leptospirosis in India is based on hospital data only . The primary reason being availability of clinical material & laboratory facilities to diagnose the disease in hospital . This is relevant for developing countries , where there is shortage of skilled manpower & finance to do community based research . It was possible for me to design guidelines for diagnosis & management of leptospirosis , based on my hospital experience for community practice also.
• asked a question related to Clinical Epidemiology
Question
I want to investigate the impact of bacterial infection, which could occur at any time of disease course, on patient prognosis. How can I choose appropriate statistical methods?
To select statistical methods without having research hypothesis first is difficult. According to your idea,  you can study
1. correlation: is having bacterial infection correlated to disease prognosis? how strong, what direction?
2. regression: could having bacterial infection predict disease prognosis?
3. differences: Is disease prognosis different between those who had infection and those who did have?
4 survival and so on.
But it all depends on your research question and hypothesis, as well as your data type. The use of parametrics or non parametrics tests depends on your data. If the assumptions were met, all could produce valid and meaningful result.
• asked a question related to Clinical Epidemiology
Question
Many researches demonstrate that increasing the ceiling height of naturally ventilated hospital wards can provide a better IAQ and allow for reducing the risk of infection transmission. It has been interpreted that increasing ceiling height increases temperature stratification which concentrates hot air above occupied space. But does anyone one the explanation of the same phenomenon regarding airborne particles and contaminants?
Dear Alia,
Smaller the room larger the ventilation rate. In any IAQ problem dilution of the pollutant concentration is done by either increasing the fresh air supply or preventing source contaminants.
About your question, if height of the room is increased, that increases the room size (by volume). It may be naturally ventilated or machine assisted, all the pollutant concentration level will automatically comes down ie. CO2 level, PM or any other airborne contaminants.
with regards,
P.Thirumal
• asked a question related to Clinical Epidemiology
Question