If the data is de-identified, then you do not need consents. Usually such kind of retrospective data study can be exempted from the IRB. However, anyway your study plan should be submitted to IRB, and have to be exempted or permitted.

The type of data and the feedback to the patients have to be considered also. If you are working, by example, upon genes or expressions of this genes in the prediction of better or worse clinical outcomes, yours results can be potentially interesting for the patients treated. Then a consent can be needed in order to ask patients if they will agree to receive these results, unexpected at the moment of the data collection. This situation has to be considered from an ethical point of view.

Overall I think it requires approval by a research ethics committee.

We had a similar challenge just last month. We asked the local ethical committee if we needed a new approval (attaching the previous application for the sample collection) for the new use of human sera. The answer was that we did have to apply for the new use of the material. We are still awaiting the answer.

I think it is ethically correct to ask your local ethical committee if there is a new use, such as a retrospective study, of previously collected data. The success of your new application will depend on what the original study informed about regarding the use of samples and collected information in their consent form (if dealing with humans).

I agree with Brian's comments. Retrospective studies of living participants are subject to IRB approval (IRB=Institutional Review Board, our ethics boards in the US). Retrospective studies simply look back in time, and may involve original, new data collection. Example: data abstracted from medical records of living individuals. Example: data obtained from previously-obtained tissue / fluid samples. Example: personal history data obtained from participants themselves. All of these require ethics committee review.

The conventional wisdom is that for "research" and publication, IRB approval is required. For internal process improvement projects, it is not. For retrospective studies, exemption should be the first course sought, particularly for studies based on administrative data, since it is usually highly impractical to obtain consent for past events. How would one obtain consent for a ten-year retrospective study of antibiotic use in community-acquired pneumonia?

I am increasingly of the contrarian but growing view that IRBs are the single greatest impediment to effective clinical research.

Although most institutions require IRB approval, consent is usually not required as the IRB will ensure that steps are taken to ensure the anonymity of the data.

The answer would be "it depends" on what you mean by retrospective.

if you are looking at database information that would generally be anonymized, then there is no way of obtaining consent, but you will need approval from the data protection authority, or committee, or whatever, and/or IRB (depends on country). Actually that is not retrospective, that is historical prospective, since the information was entered prospectively in the database, even if you go back to get it.

if you are asking patients what happened in the past (eg, what drugs they took before an accident), then obviously you need consent (even if you state that just answering the questions is an indication of consent, that is not enough, you have to ask them before whether they agree to answer). That is really retrospective.

You really need to describe exactly what you are doing or planning to do. Retrospective is usually meaningless, or at least interpreted in many different ways (and most data in a prospective study is in fact retrospective, such as post-hoc assessment of events).

Thanks for all, In fact, we are planning to conduct a study about clinical profile of bacterial meningitis in our hospital in the past five years.

in France there is no need for patient authorisation or DPA approval for access to clinical files by the hospital staff of the hospital where the patient was admitted and treated, if data is managed by the therapeutic team, for purposes of improving patient management or internal evaluations of practice. You will however need to ensure that any communication is strictly on anonymized data or better still on agggregated data only. It is especially important for case reports to avoid things like "mr A..., 55 years old, married 3 childrent, bank employee, was admitted to our department from June 12 to 25, 2005 for acute gonococcal meningitis", from which anyone can identify the subject. This might get you into trouble...

I agree with Patrick and Nicholas. In Argentina consent is usually not required as the IRB will ensure that steps are taken to ensure the anonymity and confidentiality of the data. Moreover, this is supported by the "International Ethical Guidelines for Biomedical Research Involving Human Subjects", published by PWO/WHO, Cientific Publication No.563,pp 25, 1996.

Informed consent is a process with an individual/a research participant. He /she has a right over the information given by him/her and the sample that is is going to be used for a specific purpose. If there is possible use of it in future the right to know about it rests with him whether it can be used or not. All this needs to be stated in the ICF. Has this information been given to him for future use given adequately. Now when retrospective data and sample is used, this is an additional research issue for which the individual can be traced and an IC taken. If its an old sample set where the participants cannot be traced and there is no information on its future use it would be imperative that the permission from the IRB be sought giving reasons why it is critical to use this data and the sample for further analysis.eg left over samples from an epidemic need to used to understand another related epidemic. The EC may consider the important contribution to science it may make, They may suggest using the sample for related tests , yet anonymizing the data.

Dear Mawar, are there specific and clear rules in your council to deal with such situations?

Yes the rules are very clear that permission of the IRB must be taken to reuse sample. ( See ICMR Ethical Guidelines 2006) EC has advised tracking subjects and taking a fresh ICF with the new study mentioned. Where a participant refuses that sample should not be used.

In case of using very old samples a specific reason for using the sample was given (it was in relation with an epidemic and a new emerging epidemic) and only when EC was convinced that the study would add to the larger interest ie gaining knowledge for its application the study was approved.

Although when it it easy is better to ask for the approval, consent is usually not required because it is difficult to get, specially in patients treated many years ago. What is mandatory is to take all steps to ensure the anonymity of the data.

Juan Viñas

Thanks for all, I think all of you agree on the necessity to have ethical approval and the anonymity of the data. In fact in our hospital, to approve your retrospective study by the research committee you have to have a waiver of Informed consent from the committee.

I think only approval from IRB is enough for that without any consent of the patient

Thanks Sami.

I think there is no absolute need for consent in using secondary data. Waiver is requested from the IRB. The IRB will also advise further based on the type of research.

Thanks Adamu

In my institution only approval from a research Ethics committee is needed and without any consent of the patient because it is difficult to get, specially in people studied many years ago.

consnt is not needed for retrospective study but waiver of informed consent is required by some IRBs

Thanks Mohamed and Carlos

For a retrospective study, i dont think there is need for further consent with secondary data aside the IRB and ethical committee approval. Nonetheless, i have heard of editors requesting for renew consent in field study where secondary data was employed. It was an study involving cancer patient. Thus, i think the type of research determine as well as the decision of the ethical committee.

Consent is an ethical issue. Usually it is obtained from participant persons, their guardian or sponsors, regardless of whether the study is retrospective (e.g., retrospective interview) or prospective. However, if the retrospective study is based on recorded data, then approval from owners of the data is required. In addition any ethical body or committee should approve the study design and use of data for the purpose  intended.

the patients will not be reachable and it will be tedious to get the consent. however consent from respective institution for using the data is feasible and one time process.

Is informed consent required for original articles based on retrospective study of histopathology slides.

Dear Fahmi. No, it is not necessary as long as the anonymity of the participants is respected (name, document, medical chart, etc.)

I think it is not necessary. Ethical Reserach Committee must take care of the anonymity of the participans. In fact, it is very difficult to contact all patients in a retrospective research

If the confidentiality is garanteed, due to de difficulty to obtaind retrospectivaly the Inform Consent, I think it it not necessary.

Obviously, as Sorin writes, it is necessary that the stugy is apporved by an ethical acredited committee.

Agree with Brian and Michael

yes

its required as GCP guidelines says that if we will use patient file information in publication of study results, you need to inform him, if patient is not available, you need to inform his relative

What kind of consent do you obtain for a retrospective study on the interobserver agreement on the predictive value of EEG in patients after cardiac arrest? We will assess 100 consecutive bedside EEG on such patients seen independently by 4 experts and compare with the outcome.

consent is taken in some institutes at the time of making indoor/ outdoor files of all patients as we cannot predict which patients data will be used for future research! so then question arises

is ethical committee permission necessary for the retrospective study on institute data??

according to me yes it is required!

please opine

Yes Suvarna

You are right

Consenting is taken during filing of patients, and it's more ethical to reconsent before beginning retrospective trials, to be sure that he is still remember what he was signing for .

Thanks very much

Based on the sample size and requirements and necessary of the project the it would be safer to have a consent form from the organization the data is being used. In most instances a consent form is not required unless it involves medical data of patients

Dear Fahmi, in some countries such morocco it is not required, but while submitting, the journal may ask for ethical committee approval

It is necessary and safe to obtain consent from the organization the data is being sourced from. This will protect the organization and by extension the patients ( their clients).

True, I agree Dr. Caroline