10.08.2013 Views

Het volume van chirurgische ingrepen en de impact ervan op ... - KCE

Het volume van chirurgische ingrepen en de impact ervan op ... - KCE

Het volume van chirurgische ingrepen en de impact ervan op ... - KCE

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>KCE</strong> Reports 113 Volume Outcome 15<br />

The initial i<strong>de</strong>a was to study the <strong>volume</strong> outcome relationship for 32 procedures and<br />

medical conditions. To this <strong>en</strong>d, we selected more than 400 000 hospital stays in the<br />

MCD 2004. This large amount of data, however, obliged us to limit the IMA data that<br />

were linked to MCD; although IMA has information on the use of (neo) adju<strong>van</strong>t<br />

therapy, only the <strong>de</strong>cease date of pati<strong>en</strong>ts was retrieved from the IMA database.<br />

Nevertheless, wh<strong>en</strong> we were faced with the abundance of 32 procedures and<br />

conditions, it was <strong>de</strong>ci<strong>de</strong>d to limit the analyses to 12 surgical procedures. These were<br />

chos<strong>en</strong> from 3 medical domains: four cardiovascular procedures, three orth<strong>op</strong>aedic and<br />

five oncologic procedures. The choice of procedures was done in agreem<strong>en</strong>t with<br />

Belgian Cancer Registry and internal <strong>KCE</strong> experts.<br />

12 procedures will be analysed in Belgian data:<br />

Four cardiovascular procedures: carotid <strong>en</strong>darterectomy and carotid st<strong>en</strong>ting;<br />

coronary artery bypass graft; heart valve replacem<strong>en</strong>t or repair; percutaneous coronary<br />

interv<strong>en</strong>tion.<br />

Three orth<strong>op</strong>aedic procedures: total hip replacem<strong>en</strong>t; total knee replacem<strong>en</strong>t; hip<br />

fracture surgery.<br />

Five oncologic procedures: oes<strong>op</strong>hageal cancer surgery; pancreatic cancer surgery;<br />

colon cancer surgery; breast cancer surgery and lung cancer surgery.<br />

2.3.2.2 Available outcome measures<br />

The following information is unavailable in the data:<br />

1. Mortality within a very specific time frame i.e. 30-day mortality is impossible<br />

since the IMA-AIM mortality date only m<strong>en</strong>tions month and year of <strong>de</strong>cease.<br />

As best possible approximation of the 30-day mortality, we inclu<strong>de</strong>d the<br />

mortality until the last day of the month following the procedure. For<br />

conv<strong>en</strong>i<strong>en</strong>ce, this is called the approximate 30 day-mortality. The rea<strong>de</strong>r<br />

should be aware, though, that this outcome measure covers a post-<strong>op</strong>erative<br />

window of at most 60 days. Two examples might clarify:<br />

• If a pati<strong>en</strong>t un<strong>de</strong>rgoes a CABG on January 1 st , 2004 and the data tell us<br />

that he <strong>de</strong>ceased by the <strong>en</strong>d of February 2004, we know that he lived at<br />

least 30 days but less than 60 days after the CABG.<br />

• A pati<strong>en</strong>t who has a PCI on January 31 st , 2004, who is <strong>de</strong>ceased by the<br />

<strong>en</strong>d of February 2004, will have lived at least 1 day but less than 30 days<br />

after the PCI.<br />

2. Mortality data for this project (provi<strong>de</strong>d by IMA-AIM) were only available<br />

until the <strong>en</strong>d of 2006. Consequ<strong>en</strong>tly, mortality of three years or more after a<br />

procedure (which took place in 2004) is not available. This implies that twoyear<br />

survival is the maximum feasible outcome measure for long-term<br />

mortality.<br />

3. Complication rates within a specific time frame are unavailable because the<br />

MCD admission date only specifies month and year. This makes it, for<br />

example, impossible to study the 30-day infection rate.<br />

4. Information about cancer recurr<strong>en</strong>ce is not readily available in the BCR data.<br />

Although information on the l<strong>en</strong>gth of stay and costs of hospital stays is available in the<br />

MFD, these measures will not be analysed while this study focuses on clinical <strong>en</strong>dpoints<br />

and not on resource utilization.<br />

Table 2.1 gives an overview of the procedures and their outcome variables that were<br />

selected to be analysed with Belgian data.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!