Questioning the value of e-health

 a l  i  t e  e a e  e l h  a e  d i i t a o s  e e  n  w  w e  t e  K  a i n l  e l h  e v c  s i  i  w u d  n e t  o e  h n £ 2  i l o  i  n w  o p t l  n o m t o  s s e s.

 t  a  a  e d i e  r b i g  u b r,  u  w s  e e y  e r s n a i e  f   g o a  s i t  o a d  u d t n  t e  o t a e  l t o m  t a  a e  s d  o  a a e  a i n s  s  h y  o e  h o g  p b i  a d  r v t  h a t  s s e s  r u d  h  w r d.

 u t a i  h s  t  f i  s a e  f  e v l  p i e  h s i a  t c n l g  p o e t , i c u i g  n t a i e  w r h  u d e s  f  i l o s  f  o l r  i  Q e n l n , V c o i n,  o t  A s r l a  n  N W.

 u  a c r i g  o  n v r i y  f  y n y  e e r h r , t e e  s  e y  i t e  c e t f c  v d n e  o  u p r  a  i e y  e d  e i f  h t  -h a t  p a f r s  u  t e  a e  f  e i a  e r r  a d  m r v  q a i y  f  a e.

 e l h  n o m t c  P o e s r  o a n  W s b o k  a s  h  U i e s t  i  n w  x m n n  e e t o i  p e c i i g  n  d c s o  s p o t  e h o o y  s  a t  f   t r e  e r  t d  o  e- e l h  y t m .

MIS asked Professor Westbrook about the research

• MIS: What prompted the research?



Professor Westbrook: Health systems in Australia and internationally are investing large amounts in health IT systems.




There is a general move to look at the potential benefits that health IT might bring and really we're interested in saying, 'If we're investing all this money, how can we be sure that we're actually going to accrue the benefits that are promised from these systems?'




So we're focusing on certain types of systems and really trying to look at what the benefits are that are expected and how we can measure those benefits, particularly in the area of patient safety.




Internationally we have big problems in the area of patient safety and errors and these systems are particularly thought to be good a trying to reduce errors.



• MIS: So at the moment people just assume that if you introduce these systems you'll get these benefits?



There are a lot of logical [assumptions]. For example these systems mean you go from a hand written order to a computerised order so it does mean people can read it. Logically, that should translate into fewer errors but when you look at these systems overseas, there isn't a lot of evidence to show they really do reduce the most severe types of errors.




Also, particularly in the US where a lot of the early studies have been undertaken, most of the research was carried out at hospitals that had developed their own system over, say, a decade. That's very different to commercial systems because if you build something that fits your organization, that's very different to taking an off-the-shelf product. In Australia we're primarily taking off-the-shelf-product.



• MIS: Why did you choose to look at electronic prescribing and decision support?



Professor Westbrook: Hospitals tend to put in e-health systems bit by bit. So what's happening in Australia is the first hospitals are just starting to move to electronic prescribing systems.




We also know that overseas work has been done looking at the prescribing component and whether that helps prescribing errors. But there's been very little done on whether these systems actually reduce the number of medication administration errors. We have a big study around that.



• MIS: You're half way through the three-year study, so what have you looked at so far?



One of the challenges in doing this research and seeing if these systems have made a difference is you have to have a baseline. We don't have good baseline data about what current prescribing errors are or what current medication administration error rates are.




We've now built that baseline and in the second stage of the study we're going back and measuring those error rates once these systems have been in operation for a couple of months.



• MIS: What was the baseline rate of errors?



I can't tell you at this stage because we're in the middle of analyses.



• MIS: But there were errors?



Oh yes, there are lots that occur. What we need to know is what the factors are that contribute to them and what the outcomes were. Many errors occur but they get picked up.



• MIS: Are you also looking at specific types of errors these systems might address?



One of the things we are looking at for the medication administration study is interruptions to clinical work. We're actually doing observational studies where we collect data on the number of times nurses are interrupted during an administration and we have found a high rate of interruptions occurs while nurses are administering medications.




In a published study we have shown that of all the work tasks that we've observed nurses do they are most likely to be interrupted when they're preparing and administering drugs. That's quite startling actually.




But we don't know whether it leads to errors and that's what our study will be able to say. We know generally that there are a number of studies that have shown that clinicians work in a very interruptive environment. No one has ever shown yet whether those interruptions lead to clinical errors.



• MIS: Given a lot of Australia healthcare providers are now looking at installing these sorts of systems, what sort of support or interest is there from the sector?



A lot of this work is done in partnership with the organisations that are implementing these systems and they have been incredibly supportive and really keen to identify and be able to demonstrate the impact that these systems are having.




Also, they want to be able to understand how if they might not be getting the level of benefit that they expected, how can the systems be modified to increase the benefit?




The other thing to note is we also are studying the new types of errors that these systems create. There are a few studies internationally that have demonstrated these systems generate their own errors. So if you imagine that you're a doctor and you want to order a drug and you've got a drop down menu and you get interrupted, you may accidentally click on the drug above or below the one you wanted.




What we're trying to do is get a much more systematic approach of the frequency of these types of errors so we can use that evidence to better design or modify these systems.



• MIS: Would you expect that if the research does show benefits it will spur further adoption of these technologies?



We have good evidence that these systems will produce benefits. I'm confident about that. I think there is a lot more we could know about how we could get greater benefits from these systems by doing fairly minor modifications.




There's often a great desire to just computerise existing paper-based systems and the way things are done, but really the potential of many of these systems is to help us look at different ways of providing care.

E r y  n  h  d c d  h a t  c r  a m n s r t r  w r  i  a e  h n  h  U  N t o a  H a t  S r i e  a d  t  o l  i v s  m r  t a  £1  b l i n  n  e  h s i a  i f r a i n  y t m .

I  w s   h a l n  g a b n  n m e , b t  a  m r l  r p e e t t v  o  a  l b l  h f  t w r s  p a i g  h  s f w r  p a f r s  h t  r  u e  t  m n g  p t e t  a  t e  m v  t r u h  u l c  n  p i a e  e l h  y t m  a o n  t e  o l .

A s r l a  a  i s  a r  h r  o  h a i y  r c d  o p t l  e h o o y  r j c s,  n l d n  i i i t v s  o t  h n r d  o  m l i n  o  d l a s  n  u e s a d,  i t r a , S u h  u t a i  a d  S .

B t  c o d n  t  U i e s t  o  S d e  r s a c e s,  h r  i  v r  l t l  s i n i i  e i e c  t  s p o t   w d l  h l  b l e  t a  e- e l h  l t o m  c t  h  r t  o  m d c l  r o s  n  i p o e  u l t  o  c r .

H a t  i f r a i s  r f s o  J h n a  e t r o  s y  t e  n v r i y  s  o  e a i i g  l c r n c  r s r b n  a d  e i i n  u p r  t c n l g  a  p r  o  a  h e  y a  s u y  f  -h a t  s s e s.

MIS asked Professor Westbrook about the research

• MIS: What prompted the research?



Professor Westbrook: Health systems in Australia and internationally are investing large amounts in health IT systems.




There is a general move to look at the potential benefits that health IT might bring and really we're interested in saying, 'If we're investing all this money, how can we be sure that we're actually going to accrue the benefits that are promised from these systems?'




So we're focusing on certain types of systems and really trying to look at what the benefits are that are expected and how we can measure those benefits, particularly in the area of patient safety.




Internationally we have big problems in the area of patient safety and errors and these systems are particularly thought to be good a trying to reduce errors.



• MIS: So at the moment people just assume that if you introduce these systems you'll get these benefits?



There are a lot of logical [assumptions]. For example these systems mean you go from a hand written order to a computerised order so it does mean people can read it. Logically, that should translate into fewer errors but when you look at these systems overseas, there isn't a lot of evidence to show they really do reduce the most severe types of errors.




Also, particularly in the US where a lot of the early studies have been undertaken, most of the research was carried out at hospitals that had developed their own system over, say, a decade. That's very different to commercial systems because if you build something that fits your organization, that's very different to taking an off-the-shelf product. In Australia we're primarily taking off-the-shelf-product.



• MIS: Why did you choose to look at electronic prescribing and decision support?



Professor Westbrook: Hospitals tend to put in e-health systems bit by bit. So what's happening in Australia is the first hospitals are just starting to move to electronic prescribing systems.




We also know that overseas work has been done looking at the prescribing component and whether that helps prescribing errors. But there's been very little done on whether these systems actually reduce the number of medication administration errors. We have a big study around that.



• MIS: You're half way through the three-year study, so what have you looked at so far?



One of the challenges in doing this research and seeing if these systems have made a difference is you have to have a baseline. We don't have good baseline data about what current prescribing errors are or what current medication administration error rates are.




We've now built that baseline and in the second stage of the study we're going back and measuring those error rates once these systems have been in operation for a couple of months.



• MIS: What was the baseline rate of errors?



I can't tell you at this stage because we're in the middle of analyses.



• MIS: But there were errors?



Oh yes, there are lots that occur. What we need to know is what the factors are that contribute to them and what the outcomes were. Many errors occur but they get picked up.



• MIS: Are you also looking at specific types of errors these systems might address?



One of the things we are looking at for the medication administration study is interruptions to clinical work. We're actually doing observational studies where we collect data on the number of times nurses are interrupted during an administration and we have found a high rate of interruptions occurs while nurses are administering medications.




In a published study we have shown that of all the work tasks that we've observed nurses do they are most likely to be interrupted when they're preparing and administering drugs. That's quite startling actually.




But we don't know whether it leads to errors and that's what our study will be able to say. We know generally that there are a number of studies that have shown that clinicians work in a very interruptive environment. No one has ever shown yet whether those interruptions lead to clinical errors.



• MIS: Given a lot of Australia healthcare providers are now looking at installing these sorts of systems, what sort of support or interest is there from the sector?



A lot of this work is done in partnership with the organisations that are implementing these systems and they have been incredibly supportive and really keen to identify and be able to demonstrate the impact that these systems are having.




Also, they want to be able to understand how if they might not be getting the level of benefit that they expected, how can the systems be modified to increase the benefit?




The other thing to note is we also are studying the new types of errors that these systems create. There are a few studies internationally that have demonstrated these systems generate their own errors. So if you imagine that you're a doctor and you want to order a drug and you've got a drop down menu and you get interrupted, you may accidentally click on the drug above or below the one you wanted.




What we're trying to do is get a much more systematic approach of the frequency of these types of errors so we can use that evidence to better design or modify these systems.



• MIS: Would you expect that if the research does show benefits it will spur further adoption of these technologies?



We have good evidence that these systems will produce benefits. I'm confident about that. I think there is a lot more we could know about how we could get greater benefits from these systems by doing fairly minor modifications.




There's often a great desire to just computerise existing paper-based systems and the way things are done, but really the potential of many of these systems is to help us look at different ways of providing care.