Mathematical Models to Predict Duration of Stay After Laparoscopic Nephrectomy Radical Simple or Partial

Laparoscopic nephrectomy (radical, simple or partial) is a minimally invasive procedure that is offered to patients for the treatment of renal malignancy or chronic renal dis-ease^ In our current medical climate, efficient use of resources in the postoperative care of patients undergoing these procedures is of prime importance^

Duration of hospital stay plays a critical role in determining the cost of individual surgical procedures (23)

Minimally invasive procedures offer significant cost savings in that they decrease duration of hospital stay (21,23) However, even within minimally invasive procedures, there may be differences in duration of hospital stay based on preoperative patient

TABLE 2 ■ Listing of Some of the Models Currently Available in Urology

Groupa

Model

Website downloads

Anagnostou (4)

Neural network for urologic oncology decision making

Bagli (5)

Neural network to predict sonographic outcome following pyelopltasty

godot.urol.uic.edu/~web/index.html

Frank (6)

Outcome prediction model for clear cell RCC patients

post-radical nephrectomy

Frank (7)

Post-op surveillance model for patients with surgically

treated clear cell RCC

Kattan, Eastham (8,9)

Nomograms to predict prostate cancer staging and progression

www.mskcc.rog/mskcc/html/l0088.cfm

Kattan (10)

Renal cancer nomogram to predict postoperative progression

www.mskcc.org/mskcc/ht/ml/6156.cfm

Krahn (11)

Model to estimate life expectancy in patients with localized

prostate cancer

Krongrad (12)

Neural network to predict quality of life in patients with

BPH or prostate cancer

Moul (13)

Neural network to predict pathological stage in

nonseminomatous testicular cancer

Moul (14)

Model to predict post-radical prostatectomy PSA recurrence

Mulhall (15)

Model to predict area of venous leak in men with erectile dysfunction

Niederberger

Models to predict presence of prostate cancer and post ESWL outcome

godot.urol.uic.edu/~web/index.html

Parekattil (16)

Model to detect bladder cancer based on urinary NMP-22,

uICAM-1 and MCP-1

Parekattil (17)

Model to predict the outcome and duration of ureteral or

www.uroengineering.com

renal calculous passage

Parekattil

Models to predict outcomes in male infertility surgery and IUT outcome

www.uroengineering.com

Tanthanuch (18)

Neural network to predict upper urinary tract calculi

Tewari (19)

Neural network for initial staging of prostate cancer patients

aGroups are listed alphabetically.

Abbreviations: RCC, renal cell carcinoma; ESWL, extracoporeal shock wave lithotripsy; PSA, prostate specific antigen; IUT, Intrauterine transfusion; BPH, benign

prostate hyperplasia.

parameters. The identification of patients who require a longer duration of hospital stay would better prepare ancillary staff for more efficient use of medical resources.

Preoperatively predicting duration of hospital stay for patients undergoing a laparoscopic nephrectomy (radical, simple or partial) would provide the hospital scheduling staff a planned approach to the recovery period. Such models could be incorporated into clinical pathways for postoperative management. The use of clinical pathways in postoperative management of patients after radical prostatectomy (2) and ureteroneocystostomy (1) has been shown to provide significant improvements in cost containment and medical resource utilization. The use of predictive models in clinical pathways may only further enhance these benefits.

Our goal was to design algorithms to predict the duration of hospital stay after laparoscopic renal surgery based on preoperative patient parameters.

There are a number of confounding variables involved in determining duration of hospital stay, such as the patient's expectations, baseline patient coping ability, surgeon bias, and timing of the surgery (e.g., would the patient be discharged on a weekend day vs. a workday, etc.). However, our goal was to identify if there were any identifiable preoperative patient characteristics that predisposed the patient to a longer duration of hospital stay.

Retrospective review was performed on all 392 patients (July 1997-March 2004) who underwent laparoscopic nephrectomy (simple or radical) and all 334 patients (September 1999-April 2004) who underwent laparoscopic partial nephrectomy at our institution.

Prospective testing of models was performed on all 29 patients who underwent a laparoscopic nephrectomy and all 19 patients who underwent a laparoscopic partial nephrectomy from May 2004 to July 2004 at our institution. The protocol was approved by our Institutional Review Board. Age of the patients in the radical/simple nephrectomy group ranged from 24 to 89 years (mean, 61). Age in the partial nephrectomy group ranged from 17 to 87 years (mean, 60). Racial background was not recorded.

0 0

Post a comment