Urgencies and DRGs

Flat rates based on Diagnosis Related Groups (DRGs) are being introduced in a growing number of countries to remunerate acute inpatient treatment. In doing so, it will have to be decided – among other things – whether inpatient emergency services should be remunerated separately from the DRG flat rates or as part of them. Without se­pa­rate remuneration, there may be an apprehension that wrong economic incentives with regard to emergency services could develop. To be able to decide about a se­pa­rate remuneration of emergency readiness and/or emergency treatment one must know how to categorise services, what are the costs of emergencies and how the remuneration can be deduced from them.

The presented material was collected for a study for the German Hospital Association DKG (Deutsche Krankenhausgesellschaft). Later it was published as a book of its one.1

[Table 1] Inpatient urgency remuneration systems of different countries were analysed using a scheme developed by combining aspects of cost calculation and tasks to perform.

Costs and cost centres:

1. Emergency services.
2. Emergency room.
3. Further hospital services.

Emergency services include the emergency telephone number, the emergency physicians, the ambulances, etc. They encompass the activities for the patient outside the hospital.

Emergency room include emergency beds with 24 hour access, experienced emergency nurses, physicians which are available within short time,

The further hospital services include hospital operations rooms and wards. There must be availablity of treatment which can be urgently planned and performed. There has to be skilled personal, free rooms, and adequate equiment.

Tasks:

1. Emergency readiness.
2. Emergency treatment.

By means of internet searches in the year 2008, the author collected information about different solutions of emergency patient classification systems and emergency flat rates in France, United Kingdom, New South Wales (Australia), and – additionally – about various approaches towards regulations in Switzerland, Germany, United States, Canada, and Victoria (Australia).

The information was collected with the aim to be able to suggest – with a view to deciding about the design of inpatient urgency remuneration systems – whether any increased emergency costs would justify se­pa­rate remuneration and how this could be set up.

In Switzerland, acute inpatient treatment will be remunerated from 2012 onward by the SwissDRG-System, an adapted GDRG-System. Following a law introduced at the end of 2007, the new flat rates must not contain public welfare services. Hence, emergency readiness has to be calculated and remunerated separately from DRG flat rates, independently of the number of cases.

There is a recommendation that remuneration for emergency readiness should be derived from the number of inhabitants.2

In the Swiss minimal data set, "emergency attendances" are defined as attendances of patients who are "required to be treated within 12 hours". As complement, "planned admission" can be chosen.3

In Germany, there is no se­pa­rate remuneration for emergency admissions. In principle, hospitals are ordered to participate in emergency services. Hospitals which do not participate have to expect a deduction of € 50 per case.

Emergency attendances can be coded in the German minimal data set, but no definition is given.4

In the british minimal data set for acute inpatient treatments, a distinction is made between elective and non-elective admission. Non-elective admissions encompass urgent admissions as well as maternity admissions, newborns, and transfers.5

When calculating treatment costs, differences between treatment with elective and non-elective admission where found within many HRGs.6

[Tables 3 and 4 ] I drew these differences of tariffs of all HRGs by means of a "spoke plot". A spoke plot is a kind of a bar chart. As base line, a circle line is used instead of a straight line. A spoke plot occupies only a third of the space of a conventional bar chart.7

The difference bar for each HRG is put on a imagined line starting at the circle base line going to the centre of the circle: this is the spoke line. The HRGs are grouped according to the HRG main categories (= MDCs). These are abbreviated by the letters shown at the circle border at the beginning of each MDC sector. The scale from the circle line to the centre shows the HRG tariffs (in english Pounds).

The coloured lines show the differences between the HRG tariff for non-elective and for elective admission. If the non-elective tariff is higher than the elective tariff the differences line is blue, else it is red. A green point shows that both tariffs have the same value.

It can be seen that most HRGs have blue lines i. e. higher tariffs for non-elective treatment than for elective treatment. Some of these differencies are very big. And they grewed from 2007 to 2008. But there are also some elective tariffs which are higher than non-elective tariffs (the red coloured differences). By comparing the two graphics, one can see that the some differences changed remarkably from 2007 to 2008.

I tried to estimate the redistributed sums. I calculated an averaged tariff for each HRG by weighting the two HRG tariffs with the number of urgent respectively elective cases as found in the "Hospital Episode Statistics".8 The estimated redistribution amounts to 9 % to 10 % of the total remuneration volume.

[Tables 5 and 6 ] I listed HRGs with huge differences in two tables. One table shows HRGs with higher tariffs for non-elective treatment, the other shows HRGs with higher tariffs for elective treatment.

[Tables 7 and 8] [Table 9] An A&E minimum dataset was introduced in october 2006.9 An A&E classification and tariff exists for HRG 3.5 (9 HRGs)10 and for HRG4 (11 HRGs).11 While the A&E HRGs of Version 3.5 are defined by investigation type and attendance disposal, A&E HRGs of Version 4 are defined by investigation category and dominant treatment category.

[Table 10] It has to be noticed that specialised services treatment – defined by lists of procedures und diagnoses – is paid by top-ups on the HRG tariff.12

In the british HRG based remuneration system called "Payments by Results" (PbR), different HRG flat rates are defined for elective and non-elective cases (HRGs = Healthcare Resource Groups = British DRGs). In this way, about 10 % of the total remuneration volume are redistributed. (Non-elective cases encompass not only emergencies but also births, newborns, and transfers.)

Additionally, there is a three-tier emergency tariff to remunerate for inpatient and outpatient emergency attendances. It is defined by means of about 10 emergency HRGs. 80 % of the emergency tariff is paid on the basis of the planned emergency attendances in order to cover emergency readiness. This is done regardless of the actual number of emergency attendances ("80/20 rule").

These emergency flat rates are paid for emergency admissions in addition to the non-elective HRG flat rate. 50 % of the latter are paid on the basis of the planned number of emergency admissions, and 50 % as per actual admissions ("differential tariff").

The name of the french DRG based remuneration system is "Tarification à l'activité" (T2A). Emergency treatments of admissed patients are not payed separatly from the french GHS flat rate (GHS = Groupes homogènes de séjours), but there is a lump sum called "Forfait annuel urgences" (FAU) [Table 13] to cover hospital emergency readiness. From 2003 to 2005 the amounts were slowly hightened, 2006 they were reduced a little bit, and since then (until 2008) they remained unchanged.13

An experimental emergency classification GPU ("Groupes de passage aux urgences") was developed before 1999 to classify emergency attendances.14 But it was never used broadly.

For several studies an emergency minimum data set RPU ("Résumé de passage aux urgences") is in use since 2002. A growing number of hospitals are contributing data. In 2007, data of 20 % of emergency attendances at hospitals was available.15

"Emergency attendances" are patients which enter through the emergency ward/department, but not for organisational reasons (e. g. not to control plasters and not to redo wound dressings).16

Average full cost of an emergency attendance was calculated as approximately € 140.17

An annual lump sum based on the size of the emergency ward/department is paid to remunerate emergency readiness. (The size of the emergency ward is calculated on the basis of the budgeted number of emergency attendances.)

Emergency admissions (emergencies with subsequent inpatient admission) are paid through GHS flat rates (GHS = "Groupes homogènes de séjours" = French DRG flat rates). Out­patient emergency attendances are paid at a flat rate of € 25.

[Table 15] [Table 16] To deal with emergency patients, a patient classification system named "Urgency and Disposition Groups" (UDG) was introduced in New South Wales beginned at the year 2001. The patients are classified by disposal type («subsequently admitted» or «emergency department only») and triage type according to the "Australasian Triage Scale" (ATS).18

[Table 18] 1991, a seven level classification of emergency departments was published.19

In the Australian minimal data set, the "urgency of admission" can be coded as "emergency" or "elective". An admission has to be categorised as "elective" if "the admission could be delayed by at least 24 hours". A provisional list of clinical conditions is defined for emergency admissions.20

80 % of emergency costs (for inpatient and outpatient cases) are paid by a budget for emergency readiness. To this end, the planned cases are weighted by means of emergency patient classification system UDG ("Urgency and Disposition Groups") which defines 11 patient categories. Three base rates are used according to the types of hospital. (The three types of hospital are: "general referral hospitals" or "large metropolitan districts"; "childrens"; "small metro districts" or "rural base".)

The remaining 20 % of emergency costs are paid by UDG weighted emergency flat rates. For emergency admissions, an ARDRG flat rate is paid additionally.

The main suggestions are:

1. Emergency readiness should be defined and remunerated by performance contracts.
   A bonus system could promote the attainment of certain emergency targets.
2. To be able to assess the costs of emergency treatment, all DRGs should be split as per the criterion "with/without emergency attendance".
   The concept of "emergency attendance" must therefore be defined. A medical definition would be: "Emergency attendances are attendances of patients who are required to be treated within x (e. g. 12) hours."
   If cost differences arise, these can be taken into account by applying se­pa­rate DRG weights for DRGs "with emergency attendance" and DRGs "without emergency attendance".