Using health care data for decision making - Essay Example

?Health care decision making for All Saints Hospital Funding on a per diem rate versus on a DRG basis All Saints Hospital is reputed for innovative care, introduction of advanced technologies, teaching and research. It has an inner-city location and serves a rapidly aging population. The chief source of funding for acute inpatients is private insurers and Department of Veteran’s Affairs. Together they pay for close to 90% of all inpatients and their mode of payment is based according to Diagnosis related groups (DRG). Since the hospital is in financial difficulty, its source of funding needs to be revisited and possibly an alternative manner of payment negotiated. A possible alternative payment may be per diem instead of based on DRG. Both DRG and per diem payment rate is based on International Classification of Diseases (ICD); for surgical cases per diem uses Medicare Benefits Schedule (MBS), which may not reflect hospital-relative cost (Brannen 1999, Hanning 2005). Both systems take into account average length of stay in the hospital, which is argued to be one major factor underlying financial difficulties. However, when close to one-fifth of the inpatients are over 65 years of age, average length of stay computed for all ages may not be a true representation of the case. Aged patients are expected to have other co-morbid conditions, making their length of stay longer than the national average. Introduction of a per diem basis payment may take care of this aspect of the cost. However, in case of surgical patients, the flat per diem payment rates are based on very broad classifications and it may not always address the differential costing of treating less and more severe patients. DRG payment system is more efficacious to include additional diagnoses and complications as secondary diagnoses, thereby making adjustment to the costs (Ferguson 2004). There may also be a moral dilemma for a hospital to possibly overcharge if per diem system is adopted (Jian & Guo 2009). In case of All Saints Hospital that may not be important as this is a charitable hospital with a limited number of beds and serving an inner-city population. Moreover, being a teaching and research hospital, cost per patient is more compared to a non-teaching hospital (Gottlober et el. 2001). A better strategy from the hospital’s point of view may be to combine both modes of payment. Payment according to DRG may be added on by disproportionate share adjustment and outlier adjustment methods prevalent in United States (Centers for Medicare and Medicaid Services 2011). 2. Versions of DRG In November 2008 AR-DRG V6.0 has been released. This incorporates the sixth edition of ICD-10-AM/ACHI within the structure of AR-DRG version 5.2. Several new classes have been introduced to keep the system up to date where ventilation and ventilator support is necessary (Australian Govt Dept of Health and Ageing 2009). DRG separations are based on primary diagnosis (medical) or procedures (surgical) and presence of complications and/or co-morbidities. However, age is no longer a factor for DRG classification. While in AR-DRG V5.0 E66A (Major Chest Trauma Age > 69 W CC), E66B (Major Chest Trauma Age > 69 or W CC) and E66C (Major Chest Trauma Age < 70 W/O CC) took care of age-based splits, in AR-DRG 6.0 E66A (Major Chest Trauma W Catastrophic CC), E66B (Major Chest Trauma W Severe or Moderate CC) and E66C (E66C Major Chest Trauma W/O CC) combine patients of all ages in one single classification. The Chief Information Officer’s argument that the latest version of AR-DRG uses age as a factor, leads one to believe that All Saints Hospital may not be using AR-DRG V6.0. In fact Price Waterhouse Coopers (2009) found in their review of AR-DRG system that in many Australian states and territories AR-DRG V5.2 or even V4.2 is being used. These coding systems are not capable of correctly classifying the diagnoses or procedures in many cases. All Saints Hospital introduces new technology and is known for innovative care. The coding system needs to keep up with services rendered to make it economically viable. There are two aspects of the problem. One is adapting the latest version of AR-DRG and the other is reducing chances of miscoding. Poor data, specially neglect in including additional diagnosis and failure to include complex treatment and surgical procedures increase chances of coding error. Coding errors are financially disadvantageous to hospitals (Cheng et al. 2009). If diagnosis records are kept meticulously then it is possible to reduce human error in coding by implementation of coding software. 3. Average length of stay in hospital It is argued that All Saints Hospital serves mostly an aging inner-city population and hence national (or state) average length of stay (LOS) may be a significant underestimate in this case. After adjusting for complications and co-morbidities, whether LOS has any dependence on age is not always very clear. In a study involving 23 large New York City hospitals, the correlation coefficient of LOS on age came out to be significant (Posner & Lin 1975) but within an age-group large variability in LOS was also noted. Hence whether only age would contribute towards longer utilization of hospital resources still remains an open question. The most recent version of AR-DRG system does not recognize age to be a contributing factor in deciding on classification. On the other hand, age may lead to severity of complications and other co-morbid conditions. Through proper coding these may be taken care of. Also, DRG system of payment has a facility of paying for outliers. It is recognized within the DRG system that all patients with the same DRG will not have identical LOS. LOS is a variable and will have a range. In most cases the distribution of LOS may be assumed to have a symmetrical bell-shape and a mid-value and a measure of variability (e.g. standard deviation) will be able to describe the distribution. If the shape of LOS distribution is not symmetrical then median and difference between the first and third quartiles (inter-quartile range) would accurately model the distribution. The cases that lie beyond the natural trimming points of the distribution are considered outliers (Felder 2009). DRG system adjusts for these patients whose length of hospitalization is significantly more than the average. DRG data is also used for benchmarking hospital efficiency. If it so happens that for a subset of patients for one or more DRG classification the proportion of outliers is excessively high, then payment negotiations with insurers may be made for those individual classifications. There is always a possibility of payment adjustment based on cost of living and other indices. 4. Alternative casemix All Saints Hospital serves an inner-city population with about a fifth over the age of 65. It is possible that a substantial proportion of the patients are in the hospital for sub-acute or non-acute care. There is also a large mental health unit and there are patients requiring transfer to nursing care facilities. DRG is primarily for acute care patients and for the other patients in this hospital Australian National Sub-Acute and Non-Acute Patients (AN-SNAP) funding may be more appropriate (Gordon et al. 2009, Tooth et al. 2005). AN-SNAP classification codes are more focused on patient’s functional status and/or quality of life. The classifications are, e.g., geriatric evaluation and management, psycho-geriatric and maintenance etc, which may suit a good number of in-patients. In many instances the AN-SNAP system pays per diem, which is better for cost containment for the patients. For rehabilitation patients AN-SNAP pays for per episode per day with an outlier component as necessary. For palliative care payment is per phase of the treatment sequence per diem along with an outlier component if necessary. There has been some concern that in an aged population LOS may be on the higher side and hence, DRG funding based on average cost leads to underfunding All Saints Hospital. Many such long-stay cases could be taken care of through AN-SNAP funding. Also, DRG is not the best scheme for funding treatment of psychiatric patients (Drozd et al. 2006), unless they require acute-care treatment. National Survey of Mental Health Services may be used to classify and pay for mentally ill patients. This is another group who would be requiring longer hospital stay and may well be considered outliers, if classified as DRG patients. Given the kind of patients All Saints Hospital serve, it would improve its financial condition if different types of patients may be funded through different systems. 5. Benchmarking by peer hospitals Benchmarking by hospitals belonging to the same peer group indicates efficiency on the part of hospitals. The Australian Institute of Health and Welfare analyses costs per casemix within each DRG classification (Saint et al. 2002). Hospitals with less cost per case are considered more efficient after adjusting for clinical and social factors that may have some bearing upon the cost component. Clinical factors are adjusted by comparing within DRG classifications and social factors include, but are not limited to, age of patients and location of the hospital. Other factors that may have some influence on the efficiency of a hospital are admission practices and infrastructure of the hospital, clinical care practices, nursing and medical interventions etc. No data on All Saints Hospital is provided to do any actual benchmarking. If it is contended that the discharge practice of this hospital is not efficient within the peer group, relevant data must be collected and compared. Assuming after such comparison it did turn out that All Saints is lagging behind with its peer group, then all the other four solutions proposed above may be implemented simultaneously. For sub-acute and non-acute in-patients a different funding facility needs to be negotiated. For acute care patients extra funding for longer than normal length of stay must be negotiated from the insurers. For psychiatric patients AN-SNAP funding is to be followed. Financial difficulty for All Saints Hospital cannot be tackled on one front alone. First and foremost, accurate data collection system is to be implemented and any solution to tackle financial difficulty must be data driven. Once data is collected and compared, solution is to be implemented on several fronts simultaneously. References Brannen, T. 1999, ‘DRG reimbursement cost identifier inpatient los per diem’. Healthcare Financial Management, vol. 53, no. 4, p. 42. Hanning, B. W. T. 2005, ‘Combining DRGs and per diem payments in the private sector: the equitable payment model’. Australian Health Review, vol. 29, no. 1, pp. 80-86. Ferguson, L. 2004, ‘Casemix: key issues for health managers’ in M. Clinton (ed), Management in the Australian Health Care Industry (3rd ed), pp 318-337. Sydney: Pearson Education. Jian, W. & Guo, Y. 2009, ‘Does per-diem reimbursement necessarily increase length of stay? The case of a public psychiatric hospital’. Health Economics, vol. 18, pp 97-106. Gottlober, P., Brady, T., Robinson, B. & Davis, T. 2001, ‘Medicare hospital prospective payment system: How DRG rates are calculated and updated’. Office of Inspector General, Office of Evaluations and Inspections, Region IX. Retrieved from http://oig.hhs.gov/oei/reports/oei-09-00-00200.pdf on April 21, 2011. Centers for Medicare and Medicaid Services 2011, U.S. Department of Health and Human Services. Retrieved from http://www.cms.gov/home/rsds.asp on April 21, 2011. Australian Refined Diagnosis Related Groups (AR-DRGs). (n.d.). Retrieved April 21, 2010, from http://www.health.gov.au/internet/main/publishing.nsf/Content/health-casemix-ardrg1.htm. Pricewaterhousecoopers, 2009, ‘The review of the AR-DRG classification system development process’. Retrieved April 12, 2010, from the Department of Health and Ageing website: http://www.health.gov.au/internet/main/publishing.nsf/Content/Casemix-1/$File/Final- Report-November-2009.pdf Cheng, P., Gilchrist, A., Robinson, K. M., & Paul, L. 2009, ’The risk and consequences of clinical miscoding due to inadequate medical documentation: A case study of the impact on health services funding’. Health Information Management Journal, vol. 38, pp 35-46. Posner, J.R. & Lin, H.W. 1975. ‘Effects of age on length of hospital stay in a low-income population’. Medical Care, vol. 13, no. 10, pp 855-875. Felder, S. 2009, ‘The variance of length of stay and the optimal DRG outlier payments’. International Journal of Health Care Finance and Economics, vol. 9, pp. 279-289. Gordon, R., Eagar, K., Currow, D., & Green, J. 2009. ‘Current funding and financing issues in the Australian hospice and palliative care sector’. Journal of Pain and Symptom Management, vol. 38, no. 1, pp. 68-74. Tooth, L., McKenna, K., Goh, K., & Varghese, P. 2005. ‘Length of stay, discharge destination, and functional improvement: Utility of the Australian National Subacute and Nonacute Patient casemix classification’. Stroke, vol. 36, pp 1519-1525. Drozd, E. M., Cromwell, J., Gage, B., Maier, J., Greewald, L. M., & Goldman, H. H. 2006.’Patient casemix classification for Medicare psychiatric prospective payment’. American Journal of Psychiatry, vol. 163, pp 424-732. Saint, S., Rose, J., Lichter, A.S., Forrest, S., & McMahon, L.F. 2002. ‘Shifting costs from high- cost to low- cost diagnosis- related groups?’. Evaluation and The Health Professions, vol. 25, no. 3, pp 259-269. Read More