NSW Internship preferencing and explanation

Discussion in 'Internship' started by Mana, Jan 8, 2017.

  1. Mana

    Mana Intern (UNDS MBBS) Administrator

    Likes Received:
    Trophy Points:
    Junior Medical Officer
    I've had several requests for this information so I thought I would just publish it here.

    I take no responsibility for whatever outcome may result from you using this information, nor will I update it to be current (it is current as of 2016's application for 2017 internship). If you receive an internship at a hospital you didn't want using this information, that's your responsibility for not doing research current to that year.

    This information applies mainly to NSW medical graduates in priority 1 (i.e. not internationals and not interstaters).

    How HETI's Optimised Allocation works:

    Essentially it is important that you understand the system and how it works. If you don't, then you might blindly follow what I put down here and not adjust based on more recent data that will be available by the time you are applying, and then you may then use a suboptimal list and get a preference you didn't want.

    This information is in regards to preferencing for the Optimised Allocation pathway (Direct Regional and Rural Preferential pathway positions are allocated before these, and it is a lot easier to get into the hospital network you want if it happens to be in a DRA or RPR network. If you wish to get into one of those networks, it may be a good idea to apply for those instead).

    Basically, it aims to maximise the preferences of everyone in your priority category.

    How it does so is that it allocates a score of 0 whenever it gives someone their first preference, 1 to second, 2 to third etc, and then runs millions of simulations to find the iteration(s) that result in the lowest total score throughout the state.

    What this functionally means for you is this - assume the scenario where there are hypothetically two spots available to allocate to two applicants. One is at desirable hospital #1 and the second spot is at undesirable hospital #2.

    Applicant A and B both prefer to go to Hospital #1 over #2.

    However, Applicant A has placed Hospital #1 first preference and hospital #2 3rd preference, while Applicant B has placed #1 first again but #2 in their FOURTH preference.

    The system will then look at the total score of giving A #1 and B #2 and find that it's 0 + 3 = 3.
    It will also go through the iteration of total score of giving B #1 and A #2 and find that it's 0 + 2 = 2

    Therefore for the system in this scenario it will ALWAYS allocate B to #1 and A to #2 to minimise the total score and thus maximise everyone's preference.

    Ideally, this then means that you want to list the hospitals in order of most subscribed to least subscribed (proportional to number of places), and then take your #1 preference out of that and place it on top. If you make one mistake, and place a less desirable hospital over a more desirable one in your list, the system will find that it will score less to allocate you to that network and you'll get that one - so it is absolutely essential you do your research!

    There is also a random factor in all of this - there are many iterations which will give the same lowest score, and one of these is chosen at random (using a random seed).

    Important things to read for data for your own research:

    HETI's yearly outcome report - this year's is at http://www.heti.nsw.gov.au/Global/i...ccreditation, Education and Training 2015.pdf is the 2015 one for 2016 entry but by the time you're interning they will have an updated one on the year before - use this to predict any changes in trend.

    www.internforum.com.au - started by the NSWMSC, this is *supposed* to be a website where people can swap internships. However, the data is not cleared until the next year - so - take a look at how many people are trying to swap OUT of a particular network, and take a look at how many people are trying to swap INTO a particular network. This can help guide your preferences. Collect data early in the year before it's your turn to apply or else they may clear it and you'll be missing information!

    Preference list for 2016:

    My preference list for optimised allocation went like this (explanation follows) with approximate number of spots right after it (actual numbers are listed in the document linked above from HETI and of course is subject to change each year).

    Needless to say, I received my first preference using this list.

    1. RPAH (50)
    2. RNSH (70)
    3. POW (50)
    4. St. V (50)
    5. Concord (50)
    6. St George (70)
    7. Hornsby (50)
    8. Westmead (110)
    9. Newcastle (125)
    10. Liverpool (50)
    11. Bankstown (50)
    12. Gosford (50)
    13. Wollongong (50)
    14. Nepean (50)
    15. Blacktown (50)

    I believe that this is close to the approximate popularity of hospitals in terms of number of higher preferences though of course there is no real way to confirm this. Other people have also received their first preference using similar, but slightly different, preference lists. However, these lists also don't make any of the "mistakes" I have listed below.

    If you find the hospital network you want, move it to #1, and reorder the rest according to a list that does go from most subscribed proportionally to less subscribed proportionally (to number of places offered) then you should maximise your chances. The lower on such a list that your #1 preference happens to be before you move it to #1, the better chance you have of getting it. Thus, if you wanted, say, Wollongong, if you submitted this list with Wollongong on the top, you'd almost certainly get it, whereas if you wanted the RPAH, you'd probably find that you have a lot of competition and thus stand a much smaller chance, perhaps well under 50%. Take this into account when deciding which to preference #1 - you may wish to minimise your risk of being sent somewhere you don't want to be if you place a less preferenced hospital in your #1 slot.

    This was a slight change from the year before's suggested stack - you'll notice that I have preferenced Newcastle, a hospital outside the Sydney surrounds, ahead of Liverpool, Bankstown, Nepean and Blacktown.

    The following information refers only to reasons why hospitals were over- or under-subscribed and is not a reflection of the quality of the hospitals, of the teaching, or of other experiences of those hospitals.
    More information about each of the hospital networks and possible rotations and hospitals that you can be assigned within those networks can be found in AMSA's Internship and Residency Guide - this links to the 2016 edition, but a new edition is released each year so check the latest one.

    Blacktown has long been the most undersubscribed network in NSW and it should go dead last unless you wish to go there - preferencing this above last is mistake #1.
    Nepean is the second most undersubscribed and should go second last - preferencing this above second last is mistake #2.

    If you don't do this you risk getting either of these networks (unless something drastic changes, for example, if RPAH's network and Blacktown's network merged. That'll never happen, mind you...)

    Newcastle, traditionally one of the hospitals outside of the Sydney surrounds, used to be quite unpopular. Additionally, it also was the largest network by size with 125 places. Therefore, it would be quite risky in the past to place Newcastle this high as you'd get it - however - in the report listed above, you'll find that Newcastle/John Hunter was also the most subscribed network for the rural preferential and direct regional allocation and in fact was the only oversubscribed one in DRA - so there were actually only around 20 places available after the DRA applicants came through, and thus it was quite oversubscribed and placed high.

    The other regional centres - Gosford, Wollongong - go after Bankstown as Bankstown is within the Sydney region and is thus more popular, but not after the two most undersubscribed networks.

    Now to get to the other hospitals in the list:
    These hospitals are colloquially known as the Big Four - the most subscribed hospitals in the NSW network. Large tertiary referral centres and all close to the CBD. The order doesn't really matter too much for these because as a general rule if you want one of these and you don't put it as your #1 preference then you will not get it. Some very, very rare exceptions where someone has put one of these in #1 and one of these in #2 and then has received a place at #2. If you wish to avoid this, know that the approximate popularity of this from most to least is RPAH → RNSH → POW → St Vincent's. Alternatively, you may wish to preference St. Vincent's #2 in the very slight off chance that this gives you a better chance at the big 4.

    #5,6,7,8 should not be changed unless you have a very good reason (such as new data or significant changes to numbers of intern positions at these hospitals, or if you want this as your #1 preference). These are the hospitals which people will move to #1 to mitigate their risk - by doing so they give up any chance of getting into the so called big 4, but they stand a stronger chance of getting their #1 preference than anyone placing one of the big 4 in position #1.

    5. Concord
    Has been #5 for as long as I know. This hospital is known for its very good teaching and is certainly the next most subscribed after the big 4. It's also smaller than St George in terms of numbers of places so proportionally demand is higher.

    6. St. George
    Has been #6 for as long as I know. Not too far from the CBD and is a larger tertiary referral hospital.

    7. Hornsby
    This seems like an interesting choice to put ahead of what I think is the more popular Westmead Hospital. However, because Westmead has more than double the places, proportionally Hornsby is more oversubscribed, and thus goes just ahead of it. This place is rather popular with medical students who are also parents in my experience and it's not surprising given it's proximity to a whole lot of very high scoring selective schools. Some of the people who choose to go here are also locals who went to those nearby selective schools.

    8. Westmead
    Large tertiary referral centre which also has the largest paeds and O&G facilities in Sydney. This is probably the furthest out west people are willing to go if they live near the CBD owing to the opportunities at this hospital. It's only this low on the list because it's got so many internship places.

    9. This is the second most important spot in your preference list. I have placed Newcastle there to enhance the "stack", however, most people will find that if they have stacked correctly, and still are passed over for one of the big four in preference #1, that they will get #9. Others have placed Liverpool or Bankstown here with reasonable results (i.e. still getting their first preference). Of the three though, I would put the popularity at Newcastle (from the DRAs) → Liverpool (large tertiary centre wth more opportunities than Bankstown/Campbelltown) → Bankstown and Campbelltown. I would strongly caution against putting Bankstown here unless you live very close to Bankstown. As I said before though, Newcastle is incredibly popular with the DRA cohort so if you actually want to be in Newcastle you should apply through the DRA pathway (and then if you fail to get the place there then you get a place through Optimised Allocation by stacking Newcastle #1).

    10, 11: see #9.

    12/13: I believe Gosford is more popular than Wollongong, but have found little evidence to support either hypothesis. I put Gosford #12 for this reason. Do your research.

    Re: Special consideration
    Some students apply for special consideration. As far as I know there are only two categories that qualify for special consideration - if you have a dependent (i.e. a kid, or you're the sole carer of an elderly parent), or if you need specialised medical treatment that is only available in an area (though this doesn't work well for Sydney because you can just drive from Nepean to the RPAH if you needed that treatment at the RPAH for example.)

    If you apply for special consideration, this (I think - and corresponds to my observed data) is their modus operandi:
    1. Draw a circle of radius ONE HOUR around the location where you are currently staying (or the address that you give them, which may be your parent's house in Sydney for example)
    2. Exclude all hospital networks where the primary hospital is outside that radius
    3. Remove the preferences that lie outside that radius
    4. Re-value your preferences so that first preference is 0 points, 2nd is 1, etc.

    This can actually end up working against you especially if you are trying for a inner city Sydney hospital, because you stack as above, and then those more than one hour get removed, which, if you live in the more eastern part of Sydney, removes both Newcastle AND Liverpool from your preference list - and then Bankstown becomes #9. Since it's now #9, you can now get allocated Bankstown over some other person who stacked as above but put Bankstown #11 because it will now score less. Basically, some unfortunate souls paid the Special Consideration application fee of ~$170 to get screwed into a lower preference because they didn't understand how this works. If you are applying for special consideration because of those above reasons and you absolutely cannot risk having to move out of Sydney (to say Newcastle or Gosford or Wollongong) - which is why you would be applying for special consideration in the first place, then I would strongly suggest that you consider preferencing one of St. George, Westmead, or Hornsby #1 if you don't want to get sent to Bankstown.

    However, this works rather well if you are using special consideration to remain in a regional hospital (Newcastle, Wollongong, Gosford). That said, if you preference these hospitals first, you should have no real issue obtaining them provided the rest of your preference list was well thought out.

    You'll notice on the HETI website's videos that all the speakers they got in the video for the Optimised Allocation pathway were from Royal North Shore Hospital *except* for the one person who applied for special consideration and got Wollongong. It seems very much to me that there weren't any people who applied for special consideration who got RNSH and the above I believe explains why. Here's the video just in case you wanted to check it yourself.

    Of all the people who I have asked who lived in Sydney who asked for Special Consideration and received it, my observation thus far was this:
    If you preferenced one of the Big Four #1 and got SC → you got Bankstown
    If you preferenced St George, Hornsby or Westmead and got SC → you got your #1 preference.

    Re: Swaps

    Regarding internship "swaps" as "facilitated" by the internforum.com.au page:
    I have never seen or heard of anyone successfully use this page to actually swap a preference, especially not one into a hospital I have listed in the first eight preferences. This is largely because people swapping on there aren't happy with their preference and want to swap it for a preference with people who placed that preference that they received #1. These swaps are obviously not realistic. Similar things happen on the relevant facebook group whenever it is made by the NSWMSC - lots of swaps are offered, none are accepted. If anyone from the NSWMSC wishes to show me a counter-example, I am willing to retract this statement. However, as it stands, internforum.com.au is merely a place to collect data on where people don't want to go and where they want to go.

    Swaps within NSW are made quite difficult by the fact that you cannot swap a place and then decline it (for example, for an internship in another state) - the swap is only valid if both applicants agree to the swap. For example, if applicant A had hospital X and applicant B had hospital Y, and they both agreed to swap, and the hospitals then approved the swap, then A would be at hospital Y and B would be at hospital X. However, when A then received an interstate offer which was then accepted, then they would have had to decline their position at hospital Y (as you can't hold more than one internship position for too long, due to the National Audit). This means that the swap is no longer valid, and so B now has to go back to Hospital Y, and then the position at hospital X then goes back to HETI for allocation in a future round.

    However, there is one way in which you can swap a position to someone and then decline your position that you were swapped - and that's due to this small country across the ditch called New Zealand. Basically, the National Audit looks to see if people have accepted more than one internship around Australia (but NOT New Zealand). Thus, if you have a not-very-desirable hospital network, and you manage to find someone who has a desirable hospital network AND who is declining it to accept an internship in New Zealand, then you can swap with them - you will then go to the desirable network and they will (presumably) go to the not very desirable one. They then go to New Zealand and accept that offer and start working there, and they'll never be caught by the National Audit. However, it's also a rather unethical move, as it means that the hospital then ends up with an unfilled intern position right at the very start of the working year, which then has to be filled rather last minute.
    I state this here because HETI and the National Audit really should fix this loophole but haven't.

    Swaps between states (e.g. an internship in NSW for an internship in QLD) are generally not possible and I have not seen or heard of any cases happening.

    Best of luck in obtaining your internship where you want it to be!
    Last edited: Jan 9, 2017
    umat_star, Isiyara, Benjamin and 5 others like this.
  2. WishfulThinking

    WishfulThinking New Member

    Likes Received:
    Trophy Points:
    I've been doing some research on how to stack my intern preferences and have a few intriguing questions:
    1. In the HETI report for 2016 allocations, there were 933 positions available in NSW. 44 were filled through the optimised pathway, 19 were Aboriginal Pathway, 112 were rural preference and 229 were direct regional. How were the missing 190 places allocated??
    2. Table 9 in the same report shows that there were NO positions offered through the optimised pathway for networks 2, 4, 7, 11, 12, 14 and 15 which just happen to be the seven least popular networks. Therefore, there were really only eight networks 'in play'. This just seems so unlikely - can anyone shed any light on this?
    3. If you look at the intern swap request site, there are 62 people that wanted to swap out of networks 2, 7 and 4. How did they even get these networks in the first place if there were NO positions offered there through the optimised pathway? It seems unlikely that they requested these networks through the direct regional pathway but subsequently want to swap for one of the big 4!
    Anyone who can shed any light on all of this? Thanks so much!!
  3. Mana

    Mana Intern (UNDS MBBS) Administrator

    Likes Received:
    Trophy Points:
    Junior Medical Officer
    Nope, there were 983 places available in NSW. The stats don't appear to add up if you just use simple addition; however, bear in mind these few factors:
    1. Each position can be filled by one intern working full time or two interns working part time each (i.e. the number of applicants getting a position is NOT the same as the number of positions)
    2. These results do not account for places that were declined and offered to category 2 or lower
    3. These results do not account for places that were offered in very late offer rounds

    I think these statistics were deliberately left out because of people like you who use the report to game the system. There were definitely more than 0 people who were offered a position in those listed networks.

    See answer to #2 - they clearly were offered those positions, however, HETI has left the statistic in the report out to prevent people like yourself gaming the system using those statistics.
  4. WishfulThinking

    WishfulThinking New Member

    Likes Received:
    Trophy Points:
    Hi Mana
    Thanks for your reply:
    You're right - somehow the numbers got muddled - I meant 983 but entered 933 in error and there were 433 places offered through the optimised pathway (not 44) - my apologies.
    Only 2 places were filled with part-time interns (4) so this isn't really relevant. I don't think the declined positions account for the missing places as Table 9 refers to 'positions filled' and not 'positions offered' and this is backed up with Table 3 that shows 'positions taken up' and 'positions declined' and ditto for late offers that are also included in Table 3.
    Your conclusion, however unpalatable, must be correct however - they are deliberately not including some statistics. I think it is rather poor - I would be happy for them to do this if they clearly stated that they were doing so (and why) but to make the numbers deliberately misleading makes the whole report suspicious/worthless.
    And 'for people like me' (and you judging from your opening post) that are trying to 'game the system': anyone who doesn't use every (legal/moral) tool available to get their best possible outcome is either lazy or foolish.
  5. Mana

    Mana Intern (UNDS MBBS) Administrator

    Likes Received:
    Trophy Points:
    Junior Medical Officer
    I agree that you should take all legal, morally sound methods to get your best outcome - however, gaming the system by preferencing hospitals that you don't actually want based on demand is arguably morally unsound (i.e. if you wanted to maximise everyone's preference you would go for a compromise situation rather than go for a big 4 hospital).

    That said, the reason I've posted this on a public forum is because the more people know about this, the less effective it becomes. In that sense, I'm making it harder for people to game the system in the way that you are trying to do (as other people will use this information as well, diluting out your chances at getting what you want). Ultimately I would like to see the information circulated widely enough that stacking doesn't work, or the system gets changed.

    I do somewhat agree that if you are lazy and don't do your due diligence that your outcome is still your responsibility.
    pi likes this.

Share This Page