Establishing Limited Adverse Occurrence Screening (LAOS) and GP Peer Review in small rural hospitals in the Wimmera, Victoria 1994-95

By Dr Robert H. Hall

As the coordinator of implementing the small rural hospital’s peer review project, I thought at the start that it would be impossible. Maybe Monash University School of Rural Health could link the key people and expertise to enable it to happen.

Quality assurance is an area that I have been interested in for nearly twenty years and I think we had a chance to do it if anyone was going to be able to do it. This is not something that any one person could do. It’s the result of a lot of teamwork and the result of a lot of long term relationships. I wish to credit the 70 GPs in Victoria from Bacchus Marsh through to the South Australian border who actually pioneered this activity.

Starting from the presumption that it would be impossible, I hoped we could do something worthwhile. We actually achieved a lot more than the expected outcomes.

This Quality Assurance activity came about as a consequence of an authorisation from the Regional Health Department. With casemix introduction in Victoria we had the situation where hospitals would be paid for work that they did and the Health Department was worried that the small rural hospitals might start doing work beyond their capacity to get the money to keep the hospitals going -- "Brain Surgery for Fun and Profit". I think that was the motivation from the Health Department.

The small rural hospitals in the Wimmera Area were staffed by part-time nurses and GPs.

The part-time and limited staffing, often lacking particular skills such as coronary care or orthopaedics or whatever, all impact on the quality of services provided.

The hospitals have a requirement that there be quality assurance activities and small towns which only have one two or three doctors can’t practically do peer review.

I know Dr Vicky Sheedy, from the Cunningham Centre in Toowoomba, has prepared a slide which shows Metropolitan Peer Review with 20 doctors in a row each with an endoscope up the fundamental orifice of the person in front of them. If that’s PR, for rural doctors the scope would be going down between the legs and up their own bottom.

PR is difficult to do to and for yourself!

We realise that suburban style peer review is just not practical. It’s too difficult to be your own doctor and to operate on yourself, to even examine the back of your chest, for example. In the same way, it’s not possible to actually do Peer Review on yourself.

A brief overview of the approach.

The model that we developed was a co-operative model across different towns and we organised the doctors so that they were in three or four towns in one group and three or four towns in another group and three or four towns in another. This enabled towns away from the local area to actually do the Peer Review for each town. You wouldn’t be doing it for each other, you would be actually be doing it around the triangle and therefore it was at arms length. That was the basic approach to peer review.

One difficulty with small rural hospital QA is there is no medical administration in these hospitals. There is a part-time visiting medical records administrator, there’s very limited facilities and as a consequence there is a need for a regional infrastructure to enable QA to happen. So what we did was to have limited adverse occurrence screening (LAOS) by the medical record administrators (MRAs) of all of the discharges from the hospitals that agreed to cooperate. If a record met one of the seven criteria, anonymous photocopies were made of the relevant admission. We initially were blanking out all the patient and doctor names, hiding the town from which the record came and making sure it was all as anonymous as possible. The copy was sent to a GP from another town and the GP would look through it to identify if it was a record from which lessons could be learnt, which were then discussed at a quality circle or reference group.

The thing is you can’t just go and set a system up like this because you want to.

How do you achieve active cooperation?

Let me remind you that the Honourable Ralph Hunt, when he was Commonwealth Minister for Health in 1977, told the medical profession that they had better get a Peer Review Process in place or else the Government would impose one.

So... nothing happened.

Because in fact Peer Review is not something that is able to be imposed.

We had two key factors involved: We needed a feasible method, and local champions.

Over a period of time, the idea started to interest some people and as a consequence of that it slowly became an idea in good standing (Donald Schon "Beyond the Stable State") by the late 1980’s.

In the Wimmera area of Victoria there was a doctor from a GP background who worked in the Base Hospital as Medical Administrator, Dr Alan Wolff. Dr Wolff enrolled in the University of New England and did a Masters Degree. As part of his research project he actually looked at adverse occurrence screening as a way of focusing Peer Review. His literature search showed that, in big hospitals in the US about fifty different health problems occurring on the discharge summary allowed the records which had a problem during their admission to be identified in about 70% of cases. Fifty health problems showed up about 70% of the records that had a problem.

Now, by limiting adverse occurrence screening, picking out half a dozen things that occurred and were identified during the casemix documentation when patients were being discharged, six or seven things could give you 50% (three quarters of what the much more exhaustive task produced). That’s why its called Limited Adverse Occurrence Screening (LAOS) in the system e used in the Wimmera. That system had been implemented in the base hospital in the Wimmera where you had 30 doctors and an MRA team, so they had the size where Peer Review could be implemented.

The fact that the base hospital was doing this was a mixed blessing to implementing a similar process in the small rural hospitals in the subregion. First of all it meant that the doctors in the region knew that something locally was being done and it led to the paranoid idea that if the base hospital was involved in this process they would just use that as a method of closing down all the small hospitals and grabbing all the work and getting the casemix funding for themselves.

While we had a base hospital champion to this process we found it necessary to put a Chinese wall between the base hospital and the small rural hospitals and in fact not have the base hospital people directly involved in the implementation of this process.

We needed to obtain a local champion among the small rural hospitals. In Dimboola there was a single GP, Dr John Pickering who had seen this process happening at a meeting he went to at the Wimmera Base Hospital, thought it was a good idea and tried it out in his small rural hospital and so he was able to be a champion inside the small rural hospital group.

So we have gone from 1977 when the Health Department first officially broached the idea up to 1992. Peer Review with LAOS had been trialed in the base hospital and in 1993 one of the small rural hospitals gave it a go, and at that stage the Grampians regional office said that they would provide seed funds for something to be done about spreading the process across the small rural hospitals of the Wimmera. So they offered a seeding grant to Monash University School of Rural Health.

There was a lot of concern that the process keep information confidential, while the health service managers wanted accurate knowledge to help improve local health care. Doctors and hospitals were concerned that, if any poor outcome data was actually brought to light, people might be sued. Victoria has passed legislation to not allow that to happen as long as there is an approved QA plan in place, and at the start of every meeting in this process you announce the magic words that this is a QA process covered by regulation such and such of the legislation. All identifiable information is protected and anybody who does leak any of this information is subject to $5,000 fines and the information is not allowed to be brought forward into court cases. So there is legislative protection for these processes.

We as the facilitators had to believe in the method and the processes, and then make sure that this idea was generally accepted by the senior administrators and Boards of the local hospitals.

The School of Rural Health demonstrated commitment to this process. I was across in the United States at the time and I flew back to this meeting (and I had to go back to the States again a week later for another meeting where I was presenting papers. I personally had to pay an extra airfare to come back and do this. )

There was a routine regional meeting of all the hospital administrators, which was held in early May of 1994. We utilised this information meeting arranged by the Regional Director of Health at which the Directors of Nursing and the Chief Executives and some of the other administrative people were present from most of the hospitals in the region, to present the ideas involved. That meeting said, "Yes, alright, we think this is fair enough. Try and do it."

So we then had an information meeting in the Wimmera and we invited everybody that we could possibly think of would be possibly relevant to it. Again at that meeting we discussed the ideas and we then debated what were the concerns and barriers the people identified and what we could do about them, and we adjusted and modified the approach in the light of the ideas that people brought forward.

Once we had agreement in principle to go ahead we then set up a steering committee that covered all the interest groups. We had a representative of the medical record administrators (MRAs) who needed to be involved in the process. We had several representatives of the GPs, we had a medical association (AMA) leader who had a strong knowledge of medical practice in that rural area and had been previously an executive member of the State Branch of the AMA and Chairman of the Federal Branch of the AMA and while he was too busy to be involved in more than a couple of meetings he certainly kept a watchful brief on it and could have killed the idea at any stage.

If we hadn’t had him involved then that might have stopped it.

Opinion leaders

To analogise the successful approach, if you are going to implement a project like this, it’s almost an anthropological exercise. You have to identify who the "tribes" are that you have to deal with and you have got to get agreement from all of them and you need to know who the opinion leaders are and get agreement from them. A key question was, as an outsider, how do you know who the opinion leaders are? At an International Society of Quality Assurance (ISQA) conference the year before, I had discussed this idea with Jack Lomas from Canada and he basically said "You ask people to nominate who they think the opinion leaders are, and then you go along to the nominees and say ‘Look, other people think you are an opinion leader. Are you willing to cooperate in this project’" and he said that most people, on being told that everybody else thinks that they are an opinion leader, are in fact willing to go along with the process. So we needed to have the cooperation of the opinion leaders among the GPs and some of those people don't go to meetings, some of them have other people going to the meetings and then have a chat to the person afterwards to learn what was done.

I personally became aware of the importance of opinion leaders when I was setting up a rural health unit in a small country town and I needed the cooperation of the Anglican Church. I talked with the vicar, and I went and talked to the Bishop about using some church facilities because we didn’t want to build new buildings. He said "Robert, you’ve got to go and talk to these two maiden ladies and their 80 year old mother as they are the opinion leaders in this parish." So I went and had afternoon tea with these three women. We discussed my ideas for a couple of hours, and at the end of the time they said "Robert, you can go ahead and do it. We will allow the local Anglican Church to be used for health-related activities." So I became aware of the informal power structures in rural communities that we needed to communicate with, and receive their acceptance.

Now we come efficiently identifying the medical records to be reviewed. We started with the limited adverse occurrence indicators identified by Alan Wolff, modified by the West Wimmera GP Reference Group. These are noted by the medical record administrator at the time of the separation of the patient from the hospital. The good thing is that it only takes about one minute extra for the discharge coding. Any additional activity that you want to involve in busy rural practice needs to be very efficient.

In this case we asked the medical administrators to look at all the 13,000 files from patients over eighteen months. If you add up how many hours were involved in this it turned out that it was about 200 hours right across the whole system, so you need to keep that in mind in doing it. The records which were screened positive (which turned out to be only 5%, range 1% to 12%, of them) we then photocopied, which took about 10 minutes each, and those photocopies were sent to the GP Divisional Office to the Project Officer that we’d hired and then they were distributed by certified mail to the GP Peer Reviewers in different towns and they spent on average 12 minutes reviewing them. If there were lessons to be learnt that the doctors and their patients could benefit from (about 7%) then the reviewer had to spend about five minutes writing a summary form.

These were returned to the project officer, who arranged for them to go to a Reference Panel.

Non-relevant records were then returned by the project officer to the MRA’s for shredding.

Learning the lessons

The reference panels were required to have acceptance from the GPs. We needed to have highly respected and local subregional chairpersons. We asked the Division to identify who they thought might do that, and then I would ring up and have a chat to them although in the early days I actually met personally with the doctors in every town and got a sense from the interactions who looked at who and who spoke and who didn’t and things like that during the information meetings to identify who were the leaders in this context. We got the natural leaders to be involved in this process. We didn’t come along with outside people or people who were keen but not natural leaders. Every town was represented; the towns had a back up person if one person couldn’t go. We had some problems if there was only one person in the town, which we dealt with by teleconferencing in people, which was a new innovation that I reported at the World Organisation of Family Doctors conference in Stockholm in July, 1996.

For the first year of this process we needed to have a person from the School of Rural Health or the Division who was not the chairperson of the Reference Group, but who was a facilitator of the discussions. We have now, after two years reached a stage where people know the process, they have been involved in it enough so that we don’t actually need to continue having an outside facilitator. I don’t have to drive 700 km to go to all the meetings, which I am pleased about. That took me away from my own practise and my own teaching responsibilities don’t get done.

The results

We started with 24 GPs in ten towns and it built up to 45 GPs and then 70 GPs this year with 19 towns. (One of them has just been re graded from acute hospital down to nursing home and so they are no longer in the system.) So we have 18 towns involved with 70 GPs and basically this is a slice of our state from Bacchus Marsh which is close to being an outer suburban area by now although it is about 45km from Melbourne and then right back to the South Australian border. Over an 18 month period we looked at 13,000 records and 674 met the LAOS screening criteria and 44 were reviewed by the reference panels and 106 changes were recommended to the system. They are the basic results.

The next issue is what do you do with results like that. We had feed back to the towns. Now some of the GPs were interested in having discussion about everyone of their cases that had gone to the reference panels, the issue here is that the individual wanted feedback because the individual was usually in a town on his/her own and didn’t have other people to talk to about clinical difficulties. What we were on about was "no names no pack drill", we were not about pointing fingers or doing witch hunts or doing anything which would lead to people feeling they were under attack. What we were doing was coming up with health system recommendations which would improve health care in the region. So we prepared reports which would go to each of the towns and they would go to the towns without having any one’s name attached to the problem.

We used initials, so we would have a report such as LW a 74 year old man came into hospital with a bleeding peptic ulcer from non steroidal anti-inflammatory medicine and as a consequence they needed to have a gastroscopy and a quick operation or something like that, which led to that person getting heart failure from too much fluid being run in to stop them bleeding to death.

Recommendation: when you are going to do blood transfusions you need to be careful about the cardiac status of the patient or something like that.

But those recommendations go back to the local level and are discussed by the local doctors and they work out how they apply these ideas at the local level.

As well as the GPs being involved, the hospital administrators wanted to know what was going on. The GPs were not interested in the personal details of individual patients going to the hospital administration. It was agreed at the Steering Group that we presented quarterly statistics and we identified what were the common limited adverse occurrences which were occurring and therefore the administrators were able to monitor at a statistical level but not a specific content level.

These are some of the issues that we looked at. An interesting thing about using non- steroidal anti-inflammatory drugs -- about 1 person in 1000 has a bleed as a consequence of these medications. Our doctors noted that we had about 12 who had a bleed from over 13,000 patients, which is quite normal on the statistics. On a norm-based quality assurance activity, which is commonly applied by hospitals, no one would have picked that up or said there is anything wrong with our finding. It’s just normal. But the doctors were concerned with individual cases. They were becoming aware that across all the towns most doctors had no such incidents, no one had two but once we grouped the whole sub-regional data together there was a normal distribution of these events. So we debated at least, since some of them were fairly serious and needed blood transfusions and whatever, how could we prevent them. We tried to come up with guidelines about it and we came to the conclusion that the evidence was there that "Losec" and "Cytotec" were two drugs that would potentially avoid this happening. We knew that they cost a lot of money, over $100 per month to have Losec for example. The doctors in the group said "oh well, no one is going to do anything about this". We prepared a letter to the Pharmaceutical Benefits Advisory Committee, and so did some other specialist groups and lo and behold in November 1995 "Cytotec" came on to the Pharmaceutical Benefits Scheme and the doctors couldn’t believe it! There had been a system change to which we had made some small contribution.

There had been various administrative changes to medical records as a consequence of this process and in fact from the literature and discussions with people in other countries the best way to get your medical records improved is to run a Peer Review or QA activity of any sort, doesn’t matter what it is, doesn’t matter if it does any good at all it will certainly improve the medical records. If you know that someone else is going to read your records, well then you write them better.

Now the key issue with this process is that we cannot impose a process like this. As Ralph Hunt said in 1977, you’ve got to have a Peer Review process or else we’ll impose one; but in fact he wasn’t able to do it, the Health Department has not been able to do it by imposition. It is only something that can happen with feeling unthreatened. Co-operation needed champions at the local level. Dr Alan Wolff and Dr John Pickering made that possible.

Cost efficiency

It costs about $13-15,000 per year for each of the three or four towns. The WestVic Division of GP cost running this process for about $50,000 per year right across the Grampians Health Region. If any one hospital tried to do that, it would cost them probably three to five times as much. This is a very efficient process by getting the cooperation of the GPs from all the different towns and since each of the hospitals have to have a QA process of some sort and the state can provide some funding to help with QA processes it is feasible economically.

This is a very efficient system as I mentioned: you only have to pick up on 7 or 8 outcomes which are easily identified by the medical record administrators at time of discharge compared to going through all the records and so it’s useful as a Peer Review process.

However, this activity is not sufficient for quality assurance in small rural hospitals.

One of our key problems we have found is that working with the GPs alone is insufficient. If you look at unstable angina management, for example, we found that the issue of adequate support staff became important. It wasn’t just a question of what could the GPs do in this situation to get things right, but the fact that they didn’t have the nursing staff with the skills to do the cardiac monitoring. The fact that you had good GPs but you didn’t have the right qualifications among the nurses meant that that hospital would struggle to provide that service satisfactorily. Certainly that was the view of the Cardiologist in the base hospital who wanted all the patients transferred into the base hospital and so when we were setting up guidelines we asked the local specialists and the local specialists said "Send them all to us".

We asked the Rural Doctors Association about this and they said, "In certain circumstances in fact the patient may die by the time they reach a specialist. With heart attacks, you need to be willing and able to give streptokinase promptly. If it’s given within an hour it has a much better effect and so in fact we came up with recommendations which were suited to the local level and didn’t satisfy the specialists necessarily but produced the best outcomes as a result of local discussion. But the problem of system change was the other thing that worried the GP peer reviewers behind that example. So LAOS based peer review is an efficient process but it’s not sufficient.

We have mentioned the fact the Steering Group was widely representative. We had representatives of the CEOs, the medical record administrators and the AMA and anyone else you could think of. In fact, we included a certain informal group of doctors, not just divisional official ones, but the West Wimmera Doctors’ group which is an informal group that has been meeting for about 10 years so that we had people who were willing to monitor what was going on. We could talk to those people and find out what they were thinking. We didn’t want anyone pulling the plug on this.

I remember a process when we were setting up a community health centre in the Melbourne metropolitan area about 20 years ago and I had someone ring me from the Amalgamated Metal Workers’ Union, who said "I am the editor of the AMWU paper and I would like to do an article on this community health centre development that you are involved with." I replied "Thanks very much. I am very busy this week. I can fit you in next week." A month later, I learned from the Committee of management that the whole project was "blackballed".

We just need to be aware that there are these informal influential behind-the-scenes opinion leaders, and we need to deal with them. In this peer review case, after we had one of our meetings a steering group representative who was not the key opinion leader, went back and his opinion leader asked him about it and he said, one of your cases got discussed at the meeting and we were critical of it. The opinion leader said, in effect, "That's not fair. How can you possible talk about my case when I am not there." That group of doctors really got close to saying they were not prepared to go along with this peer review activity. When we heard about this on the grapevine, we discussed who was the best person to ring the chap up and have a chat to him and say "We were teaching, demonstrating on this and we had to pick somebody’s case to talk about. It was just bad luck that it was yours. Normally, that’s not the way we are going to do it. We want to have the people present if we are discussing their case, even though it is anonymous. It was just part of the training". He accepted that that was the case and we were able to proceed, but these sort of activities are very sensitive and need to be handled sensitively and that’s why the steering group having representatives of all the key players was important.

We have got a CEO rep on the steering group who relates the concerns and dissatisfactions of the CEOs, which we take seriously, and want to work on together. Face-to-face meeting with the steering group and the CEOs will be necessary in the future to try and work out how we can deal with the system issues, the staffing issues and some other things because this quality assurance activity will always be incomplete.

Positive outcomes

I am happy to report that we basically have a successful process which has now been running for two years and it has local hospital and general practitioner support. GPs from new towns are discussing involving themselves in their own local peer review. Self-supporting funding issues will next need to be managed. And so it goes.

Further Information

Please e-mail: Dr Robert Hall