[Woman 1] These protocols, you know these new investigators, especially these multiple-center
protocols, are really getting to be a problem.
I don't know if we should try and develop some sort of checklist or something like that,
but you know, the third time in it seems like, you know, we would not have to go through
so much.
Let's just take a look at this one one more time, and see, can we get it straightened
out.
[Woman 2] Sure, it's a multi-center protocol, you've seen it before.
We didn't get enough information human subjects part of the protocol itself, and the consent
form has some difficulties, too.
[Woman 1] Was it the language difficulty primarily, or—?
[Woman 2] There's a lot of scientific terms in it, and I believe the committee wouldn't
agree that there was enough description of the type and amount of blood that's going
to be drawn.
[Woman 1] Okay.
Does it give them an idea of how many visits they have to make?
[Woman 2] I think it's weak in that area, too.
[Woman 1] Okay, alright, okay.
What we gotta do—you talked to him?
[Woman 2] I talked with him several times, he's enthusiastic, he's a young investigator,
it's the first time he's tried to go through the committee.
I think right now, he doesn't understand what they want, and he's getting angry.
[Woman 1] Okay, so he's at the hostile level now, so—?
[Woman 2] Almost.
Nice guy, I mean, he's approachable.
[Woman 1] So you think I should call him, or—?
[Woman 2] I think that would be wise.
[Woman 1] Okay, alright, okay.
Let's look at this behavioral one that's a problem.
Okay.
This investigator wants to do this video tape, and I'm really concerned about, you know,
whether the people will know that he might use it for teaching purposes.
I don't want him to be tempted to do that.
[Woman 2] I think it's gonna be difficult to explain that to this fellow.
He's already upset because he feels that the IRB is interfering with his work.
He wants to go ahead, and— [Woman 1] So you're afraid if he tells them,
that he won't be able to get people to participate, or—?
[Woman 2] I think he's afraid of that, I don't think he wants the consent form to
be any longer than it is.
[Woman 1] Okay, alright.
[Woman 2] I also think…
[Man 1: Edmund Pellegrino] I'm Edmund Pellegrino, Director of the Kennedy Institute for Ethics
at Georgetown University, and professor of medicine.
The comments you just heard reflect some of the common confusions and misunderstandings
about the functions of an IRB and the criteria an IRB uses to evaluate a protocol.
These comments are typical of the viewers of the scientific community, of the people
who direct research, and even the general public.
The IRB is carrying out an important moral and social responsibility.
Scientific research involving human volunteers is permitted by society to advance medical
knowledge for the benefit of all.
But if we are to continue to have that privilege, investigators must always stress the safety
of the volunteer, and the evaluation of the benefits gained.
The process of review, therefore, is designed first, to protect the subject; second, to
provide guidance to the design and experiment that will produce the results being sought.
The IRB process also assures society that those who are knowledgeable, but not immediately
involved, will independently review the experiment, give it approval, and provide assurances that
society's moral mandate will be fulfilled.
What we will be seeing today is the review of a specific protocol designed to investigate
the effects of strenuous exercise on blood clotting.
The subjects will be normal volunteers.
They will be subjected to a series of examinations involving the drawing of blood, underwater
weighing, exercise on a treadmill, monitoring of cardiovascular and respiratory function.
The benefit to the volunteers will be minimal; by and large, the significance of the experiment
lies in the information it will provide for our understanding of normal exercise and its
effect on blood clotting and heart disease.
Let's look now at the criteria the IRB will be using to evaluate this protocol.
We want to examine the criteria themselves, the way they actually work, and what their
purposes are.
[music interlude] First, the risks to the subject must be reasonable,
in relationship to the benefits to the subject, and the importance of the knowledge to be
obtained by the research.
The risks themselves must be kept to the minimum possible to achieve the purposes of the research.
There must be an equitable selection of subjects.
Informed consent must be obtained and documented, to safeguard the right of the subject to accept
or refuse participation in the study.
The privacy and confidentiality of the subject must be safeguarded.
There must be a continuous monitoring of the data.
In January of 1981, the Food and Drug Administration and the Department of Health and Human Services
issued the regulations that now govern the performance of research involving human volunteers,
the aim of the regulation being to protect those volunteers in the process of the clinical
investigations.
For the next few moments, let's look at these criteria in a little more detail.
First, we must look at the protocol.
It must be scientifically sound.
It must be properly designed so that the knowledge being sought will, in fact, be obtained, and
the methods will yield the knowledge that is sought, and the knowledge itself must have
an importance for understanding human health or human behavior.
We must always remember we are exposing human subjects to a certain number of risks.
We have to have some kind of a moral sanction to do that.
If we were to start out with protocol, for example, that had certain scientific weaknesses,
one that wasn't properly designed, then we'd be unnecessarily exposing people to
risks.
The Institutional Review Board has a responsibility to assure itself that, in fact, the criteria
of scientific probity have been fulfilled.
It needs evidence of that fact.
That, in itself, is also a moral requirement.
We have to look at the whole protocol from the subject's point of view as well.
If you were a subject, you would want to know if the information to be gained would be worthwhile
for yourself or for some other person.
The subject cannot give valid consent unless he feels the risks are worth running.
So, whether it's biomedical research or behavioral research, the first criterion is
exactly the same: important scientific knowledge must be obtained, and the generally accepted
canons of good scientific research must be fulfilled.
Now the second criterion.
What are the risks?
We have to admit that in every experiment involving humans, clinical and behavioral,
there is some risk of pain, discomfort, or even harm to the volunteer.
Every effort must be made to minimize risk.
Even though we have a very significant possibility of helping someone, we would have to assure
ourselves and the IRB that the investigator had taken every possible precaution to reduce
those risks to the minimum necessary to fulfill the purpose of the research.
Has the research, for example, incorporated safeguards to protect against the risks?
Insofar as possible, does the researcher intend to use procedures already indicated for the
individual subject's condition?
Selection of subjects is very important to avoid discrimination, or over-selection of
vulnerable subjects.
Here, the IRB should be interested in who is selected, and from what populations.
Are particularly the susceptible populations being chosen?
Prisoners, for example, students, older people, sick people, children, people with malignant
diseases, psychotic or emotionally disturbed persons, people for whom there are language
problems.
People who are particularly vulnerable must receive special protection.
Informed consent is the central ethical issue in experimentation.
Under no circumstances may we involve a human being in experimentation without his consent.
What do we mean by informed consent?
Informed consent means that the subject has to have a full disclosure of procedures, risks,
and benefits.
What is going to be done, for what reasons, what risks are there, what are the dangers,
the kind and the possibility of discomfort, the loss of time, perhaps even the loss of
dignity.
The benefits, the alternatives, all those circumstances that go along with being a research
subject, must also be made clear to the subject.
The investigator has the responsibility to ensure himself that the person does, in fact,
understand.
That means that we have to be dealing with competent individuals, or their legally authorized
representatives.
Persons who can perceive and process the information can make a decision which is their own, on
the basis of their own values, and express their decision clearly to us.
The decision must be free of coercion.
The individual truly must volunteer.
Throughout the entire experiment, subjects must know they have the right to discontinue,
to withdraw from the experiment at any time they wish to do so.
Once involved, subjects are sometimes afraid to withdraw.
The experiment has a certain momentum of its own; subjects need to be told that not only
do they have a right to refuse to participate, but they can withdraw at any time.
And they must be able to withdraw without being penalized for any agreed-upon fee or
associated treatment.
Society gives us a mandate to involve humans, because it's the only way we can find out
about the effects of new drugs, new treatments, or new psychiatric maneuvers of various kinds.
Those new procedures have to be tried out in some human being, somewhere, sometime.
Before we can involve someone as the subject of an investigation, we must be sure to respect
their rights as a person.
Under all circumstances, privacy and confidentiality must be stringently safeguarded.
This means that the IRB must assure itself, both from the protocol and from the questioning
of the investigators, that any information detected about the volunteers will be kept
confidential, and limited to those who are authorized to know.
And the subject must be told who it is that has access to this information.
That information might be damaging, for example, to the person's reputation, to his social
status, or even to his job.
The information itself might create anxiety for the subject.
Subject privacy and capacity to control his way of life must not, in any way, be compromised.
Throughout the entire experiment, the data must be monitored by the investigators or,
preferably, by some other group.
In the course of the experiment, it may become apparent that the treatment is so beneficial
that to continue to have controls would be doing a disservice to the controls.
Or the other way around: it may become evident that the drug is very toxic or ineffective.
Then it ought to be withdrawn before the study is finished.
In the case of behavioral research, emotional trauma to subjects may be such that the information,
even though significant, is not worth getting.
This means there must be continual feedback; every one of the criteria must be monitored
in the light of how the experiment is actually going.
Approval for an IRB is not unconditional.
These are the six criteria that form the basis upon which an IRB conducts its review.
Now, let's turn to an actual experiment.
We'll follow it from the time the investigators are preparing for IRB review, through to the
actual conduct of the experiment.
[Investigator 1] And then, if you take too much blood, it might affect the performance
of the athlete.
[Pellegrino] Now let's look at the way the research team prepares well in advance for
the IRB presentation.
[Investigator 2] … I think that's pretty clearly outlined.
But this raises a bigger issue altogether, about the consent form.
I based the consent form on, you know, on previous runs we've had, and I mean, is
it readable?
Is it understandable for the average subject who's coming in here?
[Investigator 3] Well I don't think that the average subject is going to know what
50 mLs is.
So I think if you could put it in terms of teaspoons, tablespoons, something like that,
then certainly a woman is going to be able to understand how much it is.
[Pellegrino] Here's another critical point, raised by one of the members of the research
team.
Are the terms understandable to the subject?
Great care must be taken to provide the information in terms of the educational background, the
language, and the culture of the volunteer.
This may be crucial when patients are the volunteering subjects.
They are particularly prone to confuse treatment and experiment.
Now let's look at the same protocol as it is presented to the IRB at the hospital sponsoring
the research.
Again, you'll see that the criteria serve to focus a thorough discussion.
[IRB Member 1] So, I wonder if you could start by giving us a breakdown on this proposal.
[Investigator 4] Okay, in this protocol, we're looking at how the fibrinolytic system – that
is, clot formation, clot dissolving – responds to acute exercise, and whether this response
differs in people of different conditioning levels.
Specifically, we're going to be look at women, and we're going to be looking at
highly trained women, moderately trained women, and then a group of untrained women.
We hope to do about 80 women in all, half on birth control pills and half not on pills,
since the pills seem to affect the clot formation system.
Essentially what we'll be asking the women to do is come in and take a maximal exercise
stress test, and we're going to be taking blood before and after this exercise bout.
And we'll also be monitoring several other factors – oxygen consumption, temperature
monitoring, and we'll have an EKG hooked up so that the people will be monitored continually
while they're up there.
This stress test will be under the supervision of a physician, and the physician will be
in the room conducting the test, and the physician draws the blood.
And then in addition, after that test is over, we will be weighing the subject under water
to determine her percent body fat.
[Pellegrino] These are the kinds of details that the IRB should ask for, because it must
make some judgement about whether the investigators' claims to protect the safety of the subject
are, in fact, true.
Reviewing what was written in the protocol is necessary to bring out possible contradictions.
Nothing should be left to chance.
[Investigator 4] We generally recruit subjects by posting announcements, say, within university,
or using in-house papers or in-house papers at other local universities who are not—occasionally
we've had an article in the newspaper soliciting subjects.
[IRB Member 2] Sue, in using humans in experimentations, it's very important that we completely be
aware of the benefit.
So not just the quality of the science, but whether or not it's worth—the purpose
of the experiment is worth the risk to the individual.
In this experiment, what exactly is the significance of the results you hope to obtain?
How important is clotting factors in people exposed to stress?
Is it really that clinically significant?
Health significant?
[Pellegrino] At this point, the IRB member is testing two things.
First, whether or not the design of this experiment is sufficiently rigorous, and the information
sufficiently significant, to warrant doing the experiment in the first place.
That is to say, to justify putting these subjects under the discomfort of a stress test.
The second thing is to determine whether or not there is some benefit to the subjects
from the information to be gained.
[Investigator 4] One of the underlying theories of—or, one of the current theories of how
atherosclerosis develops involves clotting, the laying down of small clots in the arteries.
And I think the feeling is that if exercise changes this clotting system, it may help
to prevent in the laying down of those clots, or maybe even help dissolving clots that have
already been laid down.
[IRB Member 1] Well, you said exercise, then, how much exercise do you—are they going
to take?
I mean, what are these highly conditioned women going to do?
[Investigator 4] Okay, the highly conditioned—well, first of all, let me say that the stress test
is what's called volitional max, which means the subject herself determines when the test
stops.
We ask the woman to go as long as she possibly can, because we are looking for a max test.
We want an indication of her maximum capacity for exercise.
But on the other hand, she's the one who determines when that point comes, assuming
we don't see any reason to stop the test.
So that it is volitional.
[IRB Member 3] When are the participants informed that they can stop the test?
[Pellegrino] Two very important points are being explored here.
The first is the assurance the subject can, at any time, ask for discontinuance of the
test.
Whether because of discomfort or simply because of anxiety.
And the second point is to examine whether or not the elements of informed consent have
indeed been provided for.
Even before the subject agreed to participate, she should know that she can withdraw at any
time.
How does the investigator respond?
[Investigator 4] Actually, she's told at least twice.
The first time is in—when I have an initial sort of interview with her, usually over the
phone, and I'm explaining the entire protocol procedure to her.
I will tell her in that that she may stop at any time.
She's essentially the one that determines when the max point occurs, and then she will
be reminded of that again, before she gets on the treadmill.
[IRB Member 4] Where is the test being conducted?
[Investigator 4] In our lab, in the university here.
[IRB Member 4] So not in the hospital?
[Investigator 4] That's right.
[IRB Member 5] What kind of equipment is available in the lab in case of an immediate emergency,
such as when you're drawing blood or during the stress test?
[Pellegrino] Here we have another repetition of the criterion of safety.
To determine what measures are available in case of a mishap.
Are there safeguards, and are they sufficient to minimize the risk to the subject?
[Investigator 4] We have a crash cart which includes a defibrillator and emergency drugs.
And the physician is present, supervising the test, and then everyone in the lab has
been trained in CPR, so we're all capable of doing emergency resuscitation.
[IRB Member 6] How much blood is going to be drawn?
And how often is it going to be drawn?
[Investigator 4] We have the subject seated in a blood-drawing chair, essentially, prior
to the test, and that's where the initial pre-exercise sample is drawn.
And then after she runs, she sat back down again, in that chair, and draw another.
[IRB Member 6] How much?
How much blood will be drawn in each sitting?
[Investigator 4] Each draw is 50 cc's, so that the total is around 150, which is about
a quarter of what you would give if you were to give a blood donation.
[IRB member 7] Somehow, the way you just said that, it came through to me as more blood
being taken than I was aware of when I read this study, and I identify from that where
if I were a subject, I think—possible that I might have missed it, too.
[Investigator 4] I generally tell the women, when I'm explaining to them, I say almost
exactly what I just said then, that it's about a quarter of the amount—the total
amount is about a quarter of what you would give if you went to give a blood donation.
So then it's— [IRB Member 7] Okay, I wonder from that, though,
if it shouldn't be – as opposed to being dependent on, you know, you could get sick
or something, someone has to take over and then it's not there.
I wonder if shouldn't be in the consent form, in that sort of clarity.
[IRB Member 8] Well I second your thought, Randy, because I just saw it as three little
blobs about that big.
[IRB Member 9] Yes, yes.
[IRB Member 8] Three times.
[Pellegrino] At this point, the IRB member is pursuing the question of a proper consent
form, and assuring himself and the board that it has been properly executed.
But over and above that, the IRB member is making a contribution to the way the consent
should be obtained.
The IRB review process is not simply a judicial one, or a legal one.
It also provides actual participation in improving protocol.
So as you can see, the IRB can be a help to the investigator, as well as to the subject.
[IRB Member 9] Who's responsible for withdrawing the blood from the participants?
[Investigator 4] In our lab, it's usually actually a physician who does the blood drawing.
[IRB Member 8] So there was a lot of concern here about women and the stress test.
There were some male studies I notice also, and was there anything adverse that came up
in those studies that would help us here?
[Investigator 4] I think there was—well, let me first say I was not here at the time
that study was done.
But I think there was one incident where a person collapsed after having run on a treadmill
during that study.
And I believe that he was able—he was resuscitated in the lab, and that we didn't have to call
in any emergency equipment.
So that's kind of encouraging, and it indicates to me that we have the capability to handle
that kind of emergency in the lab.
[Pellegrino] At this point we can see how important careful and consistent probing of
the protocol can be.
The response, in this case, indicates that one patient actually had collapsed while on
the treadmill test.
If one listens to the language of the investigator, we even hear the word "resuscitated."
Up to this point, the full extent of the danger to the subject had not yet been elaborated.
The investigator is conscientious, but the IRB process has served to uncover important
information of significance to the subject.
[IRB Member 3] So I was wondering if you could tell me, I know some of the women are going
to be taking oral contraceptives, but are you checking as to whether they're taking
any other medication?
[Investigator 4] We specifically ask and check to be sure.
We eliminate everybody who's on anything other than oral contraceptives, and that includes
even aspirin.
Well, particularly aspirin, because of its effects on the clotting system.
So we do check for that and we do eliminate subjects on any kind of medication.
[IRB Member 2] Could you tell me just what is being done to protect the confidentiality
of the participants in this?
Suppose, for example, the testing comes up with some information that the volunteer would
rather not have known.
What kind of protection is built into the system?
[Pellegrino] Here, we have an exploration of still another criterion: the protection
of the privacy of the subject, and the confidentiality of the information obtained during the investigation.
Information can be damaging to the patient if it is revealed when it should not be, or
if it is withheld when the subject should know about some unsuspected disorder.
[Investigator 4] The only people that have access to an individual's files are our
lab group and the individual themselves.
And we would not release that information without her consent.
[IRB Member 2] Are the files coded, or do they have people's names on them?
[Investigator 4] In the past, they've had people's names on them.
However, recently, we've started coding so that they are assigned numbers now.
So that the person couldn't just walk in, rifle through the files, and find the one
they were looking for.
[IRB Member 1] If, as a result of this study, you uncover some completely unexpected information
– for example, say as a result of the blood count, you estimated that the patient might
have leukemia, just to give example – what do you do about that?
[Investigator 4] Dr. Ferguson would inform the person of that finding, and then I think
recommend that they see their own physician.
Or, if they needed recommendation of someone to go to, we could supply them with that.
[IRB Member 7] I think that when people volunteer, they should know what they're going to get
out of the experiment.
And in this case, for the subjects who aren't runners or volunteering for that reason, they're
going to volunteer so that they can get cleared and say "I'm okay to go running now."
And it's unclear from this whether such persons will have grounds for being cleared
to go running, or whether they won't, and I think they should know that.
[Pellegrino] The question of benefit to the subject is an important one.
It should be inquired into closely, as it is being done here.
A subject might volunteer for an experiment simply out of a feeling of beneficence for
his fellow human beings, even if he or she receives no personal benefit.
[Investigator 4] If that is what the person's intent, in coming in to take the test, would
be for us to give the data to their physician and let them discuss with their physician
what kind of exercise they wanted to undertake based on the results of this test.
And I'd certainly hope that their physician would caution them about how reliable it is
to interpret this test as meaning that it's alright for them to start exercising.
I don't think anybody can give them 100% assurance on that.
[IRB Member 2] So do you have any personal hesitancy about being a volunteer for this
study?
[Investigator 4] No.
As a matter of fact, I was a subject in a pilot study.
So I don't have any problems with the protocol, although I realize from my own experience
in the lab, that sort of the more I do as a subject, the less any of it bothers me.
So I try to keep that in mind when I'm talking to people who aren't routinely subjects
in protocols of any kind.
No, I wouldn't have any problem taking part in this study.
[Pellegrino] There are some who would insist that the investigator always subject himself
or herself to the proposed procedure in a pilot test.
It is very reassuring that the investigator did just this, and experienced some of the
discomforts the subjects would undergo.
Having seen the IRB during part of its deliberation over this protocol, we might now move on and
look at how some of the concerns of the board are carried out during the actual experiment.
Here we have an opportunity to see the actual process, obtaining informed consent.
You can observe two key elements: the contractual or content aspects of consent – providing
the information, getting a signature – but also, the investigator is providing the subject
an opportunity to ask questions.
[Subject 1] So the form doesn't really go into the actual risks that would be involved.
I was wondering if you could tell me a little bit more about that.
[Investigator 4] Okay.
The risks range all the way from some relatively minor things, like maybe twisting your ankle
on the treadmill, losing your balance and falling, or we've even had—one woman had
an allergic reaction to the electrodes on her chest.
So they range from that kind of thing all the way up to possibility of heart attack
or some underlying heart disease showing up while you're on the treadmill.
[Pellegrino] The investigator is elaborating, not just providing information on a yes or
no basis.
The interaction allows the investigator to interject into the proceeding a moral quality
and assurance that, beyond the signature, there is indeed understanding.
Here we see the actual conduct of the experiment.
Even though consent has been obtained, the investigator reassures himself.
He checks again.
He wants to be sure he has provided all the information necessary to protect his subject.
It's conceivable that between the time the consent was given and the experiment actually
carried out, some change may have occurred.
This process of rechecking is very important to actually carrying out the protocol.
[Investigator 2] Okay, I don't think there are any problems.
I think we can get started now.
I need to get some baseline blood now at this point.
[Subject 1] Well in terms of the amount of blood you'll be drawing, do you think that
in any way would affect my performance on the treadmill, or how long I could stay on?
[Investigator 2] Well, I have no way of knowing that one way or the other, but because the
protocol is really designed to study effects of exercise on blood, I must take the blood,
and we've— [Pellegrino] Notice how important it was,
that even though the subject has given consent, new questions arise.
Reassurance is necessary.
The need to keep the subject informed is not a one-time event.
[Investigator 1] Okay.
Now the purpose of this is we want to measure your percent body fat, and fat doesn't weigh
anything underwater.
And so the heavier you are on the scale, the less percent body fat you'll be, because
we're just measuring your lean body mass on the scale.
So you want to be as heavy as possible, okay?
[Pellegrino] In this experiment, we're dealing with normal volunteers.
It's even more important to keep informing sick patients of what is happening to them.
They, after all, are more vulnerable, more dependent on the physician, and less apt to
bring up the kinds of questions that this normal subject raises with ease.
[Investigator 2] Now you've got a lot left in you.
You're doing real well.
In about 30 seconds, we're going to go up a little higher and a little faster.
It's still going to be a walk for you.
If you feel more comfortable jogging, go right ahead, but it should be still at a pace that
you can keep up with it simply by walking.
Do you feel alright?
[Investigator 4] Yeah, you're doing fine.
[Pellegrino] Here we are observing another criterion, the provision for constant monitoring
to ensure adequate safeguards for the volunteer.
Not only monitoring of the physiological data, but of the subject's responses.
This is important, so that if any difficulty should occur, the experiment could be stopped
before doing any harm to the volunteer.
[Investigator 2] Okay, why don't you go, but you've got a little bit left in you.
[Investigator 4] Hang on there, next click.
You're doing great.
[Investigator 2] Next click, go one more after that.
Want to try?
No?
[Investigator 4] Hang in there, okay?
[Investigator 2] Is that it?
[Investigator 4] Hang on and straddle if you want.
[Investigator 2] Grab the [inaudible] and just straddle it.
[Investigator 4] That's it.
Great.
Okay.
[Investigator 2] Hang on.
That's 10 points—clicks.
[Investigator 4] Okay, now what you want to do is step back over the blue cable just like
we did before.
Back down here, yeah.
Okay, great job, good job.
[Investigator 4] Yeah, good job.
Hold real still, now.
We'll get another EKG.
[Pellegrino] We've just seen how an IRB functions.
It looked at the way some of the criteria are actually applied by one IRB.
The criteria that we've reviewed are essential.
If, in our democratic society, we are to be permitted to continue to ask human beings
to participate in investigative procedures – which, in the long run, affect all of
us – if we are to do this, then we must be particularly careful to protect the rights
and the dignity of the volunteers.
The purpose of the IRB and the regulations that go with it, are not, thinking back to
the opening of our program, a hindrance to research.
They are not meant to stand in the way of good science; they are motivated by the moral
obligation that an investigator imposes on himself, when he undertakes to acquire knowledge
from other human beings, putting them by that fact at some risk, and causing them some discomfort.
The ends are worthwhile.
The means must be very carefully scrutinized.
That's why we need to know why the information is worth having, we must know that the scientists
who are doing it have put a great deal of effort into designing the experiment so that
it can be useful.
That we've done every attempt to minimize the risks, that we've taken every step to
be sure that the subject knows exactly what's going to happen, knows the risks that he or
she runs.
To be sure that the investigators continue monitoring their experiments.
These are the minimum requirements in a humane society for the continuance of something vital
to that society: the acquisition of knowledge, knowledge that can be a benefit for all of
us.
Không có nhận xét nào:
Đăng nhận xét