Background In 2005, Anywhere Community Hospital (ACH) in Florida was growing fast had a difficult problem with obstetrics: occupancy rates in its 17 labor and delivery (L&D) beds exceeded 80% and were rising. The unit was nearly always congested, and ACH was routinely referring patients elsewhere. In anticipation of adding new L&D beds, the departmental administrator estimated demand through 2020, as indicated in the first column of the table below. He forecasted 2.4% average annual growth, from 2,481 deliveries in 2006 to 3,449 in 2020. Out of necessity, the administrator felt that average length of stay (LOS) would have to be kept to 2.1 days through 2007, but since the state and national average was 2.2 days, the administrator felt that through 2015 this conservative LOS estimate was more appropriate for planning purposes. Table 1 ACH Demand and Occupancy 2006 - 2020 Estimated LOS Flow Time) 2.10 Annual Deliveries (Throughput) 2,481 2,535 2,615 2,694 2,773 2,984 3,028 3,082 3,134 3,192 3,245 3,293 3,348 3,396 3,449 Year 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2.20 Implied Occupancy* Deliveries Mean (Inventory) Per Day Census 17 84.0% 6.8 14.3 Beds 85.8% 6.9 14.6 63.0% 7.2 15.8 65.0% 7.4 16.2 66.9% 7.6 16.7 71.9% 8.2 18.0 73.0% 8.3 18.3 74.3% 8.4 18.6 25 75.6% Beds 8.6 18.9 77.0% 8.7 19.2 76.5% 8.9 19.1 77.6% 9.0 19.4 77.0% 9.2 19.3 78.2% 9.3 19.5 79.4% 9.4 19.8 Actual LOS 2.15 2.25 2.33 2.45 2.60 2.65 2.65 2.70 2.15 2.10 In 2005, ACH had 17 L&D beds. By Little's Law, estimated occupancies in 2006 and 2007 were 84% and 86%, respectively. (If throughput in 2006 was 6.8 deliveries/day and if LOS = 2.1 days, then the patient census averaged 14.3, or 84% of the 17-bed capacity.) Actual LOS through 2013 also appears to the far right of the table. It was also decided that target occupancy would range from 72% to 78%. Rates below 72% meant less than ideal amortization of fixed costs; and rates above 78% resulted in high congestion costs, diminished service levels, and too many patients leaving for other area hospitals. It was also decided that the expansion would aim to satisfy growth in patient demand through 2020. Not all goals could be satisfied simultaneously, and the hospital's leadership ultimately decided to add 8 L&D beds, bringing the total to 25 beds. The result was that through 2010 occupancy would be less than 72%, but at the same time occupancy would not exceed 78% until 2020. The leadership erred, in other words, on the side of adding too much capacity rather than too little. The leadership also announced the following L&D strategy: Manage congestion in 2006 and 2007 by working hard to reduce LOS, while referring patients elsewhere rather than allowing occupancy to rise above 81%; Add 8 new L&D beds in 2008, bringing the total to 25 licensed beds; Advertise ACH's L&D unit aggressively from 2008-2010 to boost demand by 5%/year; Beginning no later than 2015, initiate efforts to reduce mean LOS from 2.2 days to no more than 2.0 by 2020 (but use a conservative 2.1 day LOS estimate for planning); and Plan to add more L&D beds shortly after 2020. Regrettably, ACH's plans quickly went awry. As expected, 2006 was a difficult year, because its implied occupancy was 84%, and many patients had to be referred elsewhere. This was followed by a truly dreadful 2007. The hospital had resolved to reduce LOS to 2.1 days, but the unit was horribly congested, and LOS actually increased to 2.25 days. The chair of OB protested that LOS had also risen statewide, and for reasons beyond physicians' control. In any event, she said, the LOS increase was just 0.15 days. She asserted that higher LOS was not contributing to congestion. When new beds were added in 2008, ACH enjoyed two years of calm. Capacity had risen nearly 50%, from 17 beds to 25, so there was ample room even though LOS rose to 2.33 days. But by 2010, ACH seemed already to have outgrown its expansion, and by late 2012 a crisis atmosphere prevailed year round. Thereafter, ACH's L&D clinical operations were poor and its finances were disastrous, as mean LOS ballooned to 2.70 days and total deliveries fell far short of plans. The OB chair pointed out that the unit's LOS was no worse than state and national averages, which were also rising, but the hospital's CEO was unconvinced. How could L&D have added so much capacity in 2008 only to find that it was overwhelmed just a few years later? Analysis of Anywhere Community Hospital Indeed, for several reasons ACH's problems were actually far worse by 2013 than before the 8-bed expansion. For example, average LOS increased significantly from 2006-2013, but higher variability in Los also caused problems. In 2008, routine vaginal deliveries accounted for roughly two thirds of ACH's cases; but by 2013, they were barely half. (The table below shows that ACH's experience mirrored statewide trends.) Cesarean sections increased from one quarter to one third of all deliveries. Table 2 State of Florida Total Deliveries, 2006-2007 and 2012-2013 (Source: Agency for Health Care Administration) 2-Year Count 2006-2007 2012 - 2013 % of Mean 2-Year % of Total LOS Charges Count Total LOS Charges DRG Description 370 CESAREAN SECTION WOC 26,402 7.4% 4.3 $10,264 | 29,354 7.5% 4.4 $16,275 371 CESAREAN SECTION W/O CC 56,122 15.6% 3.1 $7,690 93,932 24.0% 3.2 $12,030 VAGINAL DELIVERY W 372 COMPLICATING DIAGNOSES 32,577 9.1% 2.5 $5,301||31,799 81% 2.8 $8,615 373 VAGINAL DELIVERY W/O COMPLICATING DIAGNOSES 220,666 61.4% 1.8 $3,814221,373 56.6% 2.2 $6,505 374 VAGINAL DELIVERY W 23,245 6.5% 2.1 STERILIZATION &/OR D&C $6,730 14,752 3.8% 2.4 $11,312 VAGINAL DELIVERY W O.R. PROC 375 179 EXCEPT STERIL &/OR D&C 0.0% 3.8 $12,931 162 0.0% 4.2 $25,435 359,191 100% 2.2 $5,222 391,372 100% 2.7 $8,924 These changes markedly increased the standard deviation of LOS. As Module 10 explains, increased variability adds to ACH's congestion and reduces available capacity. Moreover, the patients that ACH lost to other area hospitals were disproportionately well-insured. By 2013, it was therefore apparent to ACH's leadership that the total number of deliveries in 2020 would fall 25-32% short of their original plans; and that the revenue shortfall was expected to be as much as 40%! Keeping in mind that the bulk of the costs of providing inpatient care are fixed, and that half of all variable costs are tied to nursing (which was also a source of rapidly increasing costs), ACH had no recourse but to raise its L&D charges to third-party payers at a rate of nearly 10% per year. Throughout this time the relentless rise in costs and charges was a mystery to physicians providing care at the bedside. Of course, they knew that they were doing more C-sections, that nursing wages were rising, that LOS had "crept up," and that the unit had more and fancier beds. They had no idea, though, how these could combine to force a doubling of charges from 2006 to 2013, or how their actions had contributed to the increases. On the contrary, physicians could point to a variety of changes that had helped to reduce costs. Everyone was frustrated and feeling helpless. Questions 1. One of ACH's problems is that its mean LOS increased from 2.15 days in 2006 to 2.7 days in 2013. What is the % change in throughput? 2. Beginning in 2013, ACH began arranging for another increase in L&D capacity. The hospital's leaders had concluded that the higher average Los was permanent, and that demand would continue to grow as projected. Planners assumed an average 2.7 day LOS for the long run and aimed to be able to provide 3,650 deliveries per year (10 per day) with 77% average occupancy. How many L&D beds should ACH have added? 3. The OB chair believes that with better scheduling various process improvements, incremental equipment, and other modest changes, her group could reduce LOS from 2.7 days to 2.5 days and maintain 81% mean occupancy without any more congestion than under the status quo. The cost of implementing all of the chair's recommendations is $85,000. a. How many fewer beds would be needed in this case to enable 3,650 deliveries per year? b. If the chair's recommendations had come after the L&D beds in question 1 were added, how many more deliveries (beyond 3,650 per year) could ACH provide? Background In 2005, Anywhere Community Hospital (ACH) in Florida was growing fast had a difficult problem with obstetrics: occupancy rates in its 17 labor and delivery (L&D) beds exceeded 80% and were rising. The unit was nearly always congested, and ACH was routinely referring patients elsewhere. In anticipation of adding new L&D beds, the departmental administrator estimated demand through 2020, as indicated in the first column of the table below. He forecasted 2.4% average annual growth, from 2,481 deliveries in 2006 to 3,449 in 2020. Out of necessity, the administrator felt that average length of stay (LOS) would have to be kept to 2.1 days through 2007, but since the state and national average was 2.2 days, the administrator felt that through 2015 this conservative LOS estimate was more appropriate for planning purposes. Table 1 ACH Demand and Occupancy 2006 - 2020 Estimated LOS Flow Time) 2.10 Annual Deliveries (Throughput) 2,481 2,535 2,615 2,694 2,773 2,984 3,028 3,082 3,134 3,192 3,245 3,293 3,348 3,396 3,449 Year 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2.20 Implied Occupancy* Deliveries Mean (Inventory) Per Day Census 17 84.0% 6.8 14.3 Beds 85.8% 6.9 14.6 63.0% 7.2 15.8 65.0% 7.4 16.2 66.9% 7.6 16.7 71.9% 8.2 18.0 73.0% 8.3 18.3 74.3% 8.4 18.6 25 75.6% Beds 8.6 18.9 77.0% 8.7 19.2 76.5% 8.9 19.1 77.6% 9.0 19.4 77.0% 9.2 19.3 78.2% 9.3 19.5 79.4% 9.4 19.8 Actual LOS 2.15 2.25 2.33 2.45 2.60 2.65 2.65 2.70 2.15 2.10 In 2005, ACH had 17 L&D beds. By Little's Law, estimated occupancies in 2006 and 2007 were 84% and 86%, respectively. (If throughput in 2006 was 6.8 deliveries/day and if LOS = 2.1 days, then the patient census averaged 14.3, or 84% of the 17-bed capacity.) Actual LOS through 2013 also appears to the far right of the table. It was also decided that target occupancy would range from 72% to 78%. Rates below 72% meant less than ideal amortization of fixed costs; and rates above 78% resulted in high congestion costs, diminished service levels, and too many patients leaving for other area hospitals. It was also decided that the expansion would aim to satisfy growth in patient demand through 2020. Not all goals could be satisfied simultaneously, and the hospital's leadership ultimately decided to add 8 L&D beds, bringing the total to 25 beds. The result was that through 2010 occupancy would be less than 72%, but at the same time occupancy would not exceed 78% until 2020. The leadership erred, in other words, on the side of adding too much capacity rather than too little. The leadership also announced the following L&D strategy: Manage congestion in 2006 and 2007 by working hard to reduce LOS, while referring patients elsewhere rather than allowing occupancy to rise above 81%; Add 8 new L&D beds in 2008, bringing the total to 25 licensed beds; Advertise ACH's L&D unit aggressively from 2008-2010 to boost demand by 5%/year; Beginning no later than 2015, initiate efforts to reduce mean LOS from 2.2 days to no more than 2.0 by 2020 (but use a conservative 2.1 day LOS estimate for planning); and Plan to add more L&D beds shortly after 2020. Regrettably, ACH's plans quickly went awry. As expected, 2006 was a difficult year, because its implied occupancy was 84%, and many patients had to be referred elsewhere. This was followed by a truly dreadful 2007. The hospital had resolved to reduce LOS to 2.1 days, but the unit was horribly congested, and LOS actually increased to 2.25 days. The chair of OB protested that LOS had also risen statewide, and for reasons beyond physicians' control. In any event, she said, the LOS increase was just 0.15 days. She asserted that higher LOS was not contributing to congestion. When new beds were added in 2008, ACH enjoyed two years of calm. Capacity had risen nearly 50%, from 17 beds to 25, so there was ample room even though LOS rose to 2.33 days. But by 2010, ACH seemed already to have outgrown its expansion, and by late 2012 a crisis atmosphere prevailed year round. Thereafter, ACH's L&D clinical operations were poor and its finances were disastrous, as mean LOS ballooned to 2.70 days and total deliveries fell far short of plans. The OB chair pointed out that the unit's LOS was no worse than state and national averages, which were also rising, but the hospital's CEO was unconvinced. How could L&D have added so much capacity in 2008 only to find that it was overwhelmed just a few years later? Analysis of Anywhere Community Hospital Indeed, for several reasons ACH's problems were actually far worse by 2013 than before the 8-bed expansion. For example, average LOS increased significantly from 2006-2013, but higher variability in Los also caused problems. In 2008, routine vaginal deliveries accounted for roughly two thirds of ACH's cases; but by 2013, they were barely half. (The table below shows that ACH's experience mirrored statewide trends.) Cesarean sections increased from one quarter to one third of all deliveries. Table 2 State of Florida Total Deliveries, 2006-2007 and 2012-2013 (Source: Agency for Health Care Administration) 2-Year Count 2006-2007 2012 - 2013 % of Mean 2-Year % of Total LOS Charges Count Total LOS Charges DRG Description 370 CESAREAN SECTION WOC 26,402 7.4% 4.3 $10,264 | 29,354 7.5% 4.4 $16,275 371 CESAREAN SECTION W/O CC 56,122 15.6% 3.1 $7,690 93,932 24.0% 3.2 $12,030 VAGINAL DELIVERY W 372 COMPLICATING DIAGNOSES 32,577 9.1% 2.5 $5,301||31,799 81% 2.8 $8,615 373 VAGINAL DELIVERY W/O COMPLICATING DIAGNOSES 220,666 61.4% 1.8 $3,814221,373 56.6% 2.2 $6,505 374 VAGINAL DELIVERY W 23,245 6.5% 2.1 STERILIZATION &/OR D&C $6,730 14,752 3.8% 2.4 $11,312 VAGINAL DELIVERY W O.R. PROC 375 179 EXCEPT STERIL &/OR D&C 0.0% 3.8 $12,931 162 0.0% 4.2 $25,435 359,191 100% 2.2 $5,222 391,372 100% 2.7 $8,924 These changes markedly increased the standard deviation of LOS. As Module 10 explains, increased variability adds to ACH's congestion and reduces available capacity. Moreover, the patients that ACH lost to other area hospitals were disproportionately well-insured. By 2013, it was therefore apparent to ACH's leadership that the total number of deliveries in 2020 would fall 25-32% short of their original plans; and that the revenue shortfall was expected to be as much as 40%! Keeping in mind that the bulk of the costs of providing inpatient care are fixed, and that half of all variable costs are tied to nursing (which was also a source of rapidly increasing costs), ACH had no recourse but to raise its L&D charges to third-party payers at a rate of nearly 10% per year. Throughout this time the relentless rise in costs and charges was a mystery to physicians providing care at the bedside. Of course, they knew that they were doing more C-sections, that nursing wages were rising, that LOS had "crept up," and that the unit had more and fancier beds. They had no idea, though, how these could combine to force a doubling of charges from 2006 to 2013, or how their actions had contributed to the increases. On the contrary, physicians could point to a variety of changes that had helped to reduce costs. Everyone was frustrated and feeling helpless. Questions 1. One of ACH's problems is that its mean LOS increased from 2.15 days in 2006 to 2.7 days in 2013. What is the % change in throughput? 2. Beginning in 2013, ACH began arranging for another increase in L&D capacity. The hospital's leaders had concluded that the higher average Los was permanent, and that demand would continue to grow as projected. Planners assumed an average 2.7 day LOS for the long run and aimed to be able to provide 3,650 deliveries per year (10 per day) with 77% average occupancy. How many L&D beds should ACH have added? 3. The OB chair believes that with better scheduling various process improvements, incremental equipment, and other modest changes, her group could reduce LOS from 2.7 days to 2.5 days and maintain 81% mean occupancy without any more congestion than under the status quo. The cost of implementing all of the chair's recommendations is $85,000. a. How many fewer beds would be needed in this case to enable 3,650 deliveries per year? b. If the chair's recommendations had come after the L&D beds in question 1 were added, how many more deliveries (beyond 3,650 per year) could ACH provide