Functional Quadriplegia: A Nationwide Matched Study of Trends in Hospital Resource Utilization and Associated Comorbidities

Quadriplegia is a debilitating comorbidity in the hospital setting. It entails a partial to complete motor/sensory loss of both the upper and lower limbs and torso. It may also involve autonomic, bladder, bowel, and sexual functions. 

During acute encounters with patients with quadriplegia, health care providers can make an almost reflex judgment of some traumatic, ischemic, hemorrhagic, or neoplastic etiology involving the brain or spinal cord and thoroughly investigate. What does not happen so reflexively is to recognize and assign the same diagnosis to patients with comorbid conditions, such as advanced neurodegenerative disorders, congenital abnormalities, or musculoskeletal diseases, that produce a similar degree of physical incapacitation. In such cases, quadriplegia is likely to go undiagnosed and undocumented. 

On October 1, 2008, the Centers for Medicare & Medicaid Services (CMS) introduced the International Classification for Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) billable code for functional quadriplegia (FQ): 780.72.^1,2 In both ICD-9-CM and the International Classification for Diseases, Tenth Revision, Clinical Modification (ICD-10-CM), FQ is defined as “complete immobility due to severe physical disability or frailty” and excludes hysterical paralysis (300.11), immobility syndrome (728.3), neurologic quadriplegia (344.00-344.09), and quadriplegia not otherwise specified (344.00).³ Simply put, FQ is not associated with a neurologic deficit or injury and should not be coded in the presence of neurological quadriplegia.

Patients with quadriplegia, regardless of the etiology, require the highest level of care and rely on others to assist them with activities of daily living. Physical impairment may limit their ability to communicate, ambulate, bathe, toilet, and eat. Such patients are prone to developing osteoporosis, malnutrition, pressure ulcers, aspiration events, and flexion contractures and may need advanced feeding tubes and urinary catheterizations.^4,5 These factors can impact their length of stay (LOS) in the hospital and increase their risk of developing further complications, even when they are admitted for issues unrelated to quadriplegia. CMS designated the FQ diagnosis a major complication or comorbidity (MCC) in a majority of its Medicare-Severity Diagnosis-Related Groups (MS-DRGs). The designation is accompanied by a significant increase (up to triple) of inpatient payments compared with an MS-DRG without a complication or comorbidity (CC) or MCC. Caring for these patients requires a team approach from nursing, physical therapy, occupational therapy, speech therapy, wound care, and physician staff.

Data regarding the incidence of FQ documentation and its association with diseases and hospital resource utilization is lacking. In this study, we examined various trends of documented FQ and variables among this patient population, FQ incidence, MS-DRGs most commonly associated with FQ, and FQ’s association with LOS, hospitalization charges, and costs.

Methods

Data Source and Study Design

We performed a retrospective observational cohort study using the Healthcare Cost and Utilization Project (HCUP) National (Nationwide) Inpatient Sample (NIS) database, sponsored by the Agency for Healthcare Research and Quality (AHRQ).⁶ The NIS is the largest publicly available database in the United States and includes a 20% sample of all inpatient admissions at 1000 hospitals in more than 40 states. The sample averages 7 million patient discharges annually, with the sampling frame representing more than 95% of the US population.⁷ Details on the design and validity of the NIS have been described previously.⁸ Institutional Review Board approval was not required for this study because the data is deidentified and publicly available.

Variables

We used the ICD-9-CM code 780.72 to identify discharges with documented FQ (principal and all listed diagnoses) from 2008 to 2014. Charlson Comorbidity Index (CCI) was calculated for each discharge.^9,10 Hospitalization costs were estimated as described by HCUP with inflation correction for year 2014.^11,12 In all NIS-based studies, costs are estimated using cost-to-charge ratios tables offered by HCUP. Further corrections were made to adjust the weights for any missing cost ratios before any inflation adjustment.¹¹

Relative weights of MS-DRGs were obtained separately for every year (2008 through 2014) and linked to the DRG variable from NIS, which indicates the MS-DRG assigned for each discharge.^13-19

Statistical Analysis

We analyzed national estimates of trends using NIS weights for the annual prevalence of FQ as well as for mortality, hospitalization cost, and LOS among documented FQ discharges, using mixed-effects linear regression while adjusting for age, CCI, and hospital-level variables. Additional trends among hospital- and patient-level characteristics among discharges with documented FQ were assessed with univariate linear regression in GraphPad Prism 7 (GraphPad Software, La Jolla, California). MS-DRG relative weights were weighted according to their prevalence (percentage) within each calendar year and summed to produce mean-weighted overall relative weights. These weights were univariately linearly regressed against increasing years to assess trend. Since 2008 was the first year FQ was used, we detected a disproportionate underdocumentation (ie, low sensitivity) compared with subsequent years. Therefore, trend calculations included only 2009 through 2014. Additional mixed-effects multivariable regressions were conducted among the FQ population to estimate the association between various discharge destinations and LOS/costs using accelerated failure time and log-adjusted linear regression models, respectively.

To estimate the independent effect of FQ on LOS outcomes, hospitalization costs, and hospital charges, we excluded admissions with procedures or LOS equal to zero and used a 2% random sample of non-FQ discharges from each year. The resulting set of discharges (n=309,269) were matched per age group, CCI, hospital identification number, calendar year, and gender with an FQ discharge (n=5511).²⁰ A double-robust regression was used with age and CCI variables added as confounders in order to adjust for any residual variability, together with FQ (and LOS, where the outcome was cost and charges) to estimate the percent change attributed to FQ alone. After matching, 224 matching clusters for 758 observations were created for the hospital charges and LOS models, and 337 matching clusters for 1300 observations for the hospitalization costs model. P<.05 were considered statistically significant. All analyses were performed in Stata MP 14.2 (College Station, Texas) unless otherwise stated.

Results

Means and Trends

In this 7-year NIS sample, 71,906 patient discharges with a FQ diagnosis were identified (Supplementary Table 1 - at the bottom of this article). The incidence of FQ rose exponentially (R²=0.99) between 2009 and 2014 (Figure 1), exhibiting a much closer fit (P<.001) than a linear increase (R²=0.91). 

Among patients with FQ, the mean age was 71.4 years without a significant trend. Mean CCI was 1.87 with an increasing trend (1.45 in 2009 to 2.06 in 2014; P<.001). Mean LOS was 9.1 days without significant trend. Mean hospitalization costs were $19,138 without any significant trend. Women comprised 57% of patients with FQ, which remained stable over the years. Medicare was the most common payer (81.8%). The majority of patients with FQ were white (59.1%), with a decreasing trend, while among blacks and Hispanics the trend rose. 

Comorbidities that showed an increasing trend included pulmonary hypertension (P=.014), electrolyte imbalance (P=.013), obesity (P=.008), renal failure (P=.001), and arterial hypertension (P=.020). There were no significant trends in hospital characteristics. The most prevalent comorbidities in patients with FQ were arterial hypertension, electrolyte imbalance, diabetes mellitus, chronic lung diseases (chronic obstructive pulmonary disease/bronchiectasis), congestive heart failure, and renal failure. Ten percent of patients were obese.

MS-DRG & FQ Discharges (Table 1) 

MS-DRG relative weights among FQ discharges increased (P<.001) from 2.094 in 2009 to 2.354 in 2014 (Figure 2). This increase followed a linear trend (R² =.9236) and represented a growth of 12.4% over the 6-year period. Overall, the top 30 MS-DRGs represented approximately 30% of all FQ discharges from 2008 to 2014. Of note, 88.8% included MCCs, 0.2% included CCs, and 11% had no MCCs or CCs.

In 2009, 83.2% of all MS-DRGs included MCC. From 2010 to 2014, the prevalence ranged from 88.3% to 90.1%. Notably, sepsis/severe sepsis was the most common MS-DRG, accounting for 1 of every 5 discharges in patients with FQ. 

Of all 271 unique DRGs identified, there was high prevalence of infectious disease DRGs (40%). On the contrary, the FQ MS-DRGs (052 and 053) made up only 0.3% of the FQ population, suggesting that FQ is rarely the primary discharge diagnosis in FQ patients.

Discharge and Associated Hospital Outcomes in the FQ Population

Multivariable regression analysis of the FQ subset (n=13,107) on LOS revealed a 29% increased LOS (P<.001) in patients discharged to a skilled nursing facility (SNF), intermediate care facility, long-term acute care hospital (LTACH), inpatient psychiatric facility, or inpatient rehabilitation facility compared with home (Table 2). This was significantly higher (P<.001, Bonferroni-adjusted contrast comparison between coefficients) when compared with both transfer to another acute care facility (8% increase in LOS; P<.001) and discharge to home health care (20% increase in LOS; P<.001). Analysis of hospitalization costs (n=11,895) revealed patients with FQ discharged to a SNF, intermediate care facility, LTACH, inpatient psychiatric facility, or inpatient rehabilitation facility had, on average, 9.1% higher hospital costs (P<.001) compared with discharge home.

Effect of FQ on Outcomes

Analysis of 224 matched clusters (758 matched observations) revealed a 41% increase in LOS in patients with FQ (P<.001). In the same matched sample, hospital charges were 17% higher in patients with FQ (P=.025) even after correcting for LOS (Table 3). These percentages reflect an average hospital stay 1.9 days longer and hospital charges $4079 higher among the matched sample. Likewise, hospitalization costs were 18.3% higher in patients with FQ among 337 matched clusters (1300 observations). In absolute terms, this reflects an average increase of approximately $1353 among this matched sample to $3484 in the whole population with FQ exclusively attributable to FQ.

Discussion

In this study, we observed an exponential increase of documented FQ between 2009 and 2014. Furthermore, we observed the population of patients with FQ was comprised mostly of older adults. Over time, the number of comorbidities in the population increased. This was accompanied by an increase in the mean MS-DRG relative weights and high use of the MCC modifier. Lastly, matched analyses revealed an isolated effect of FQ on LOS, hospital charges, and costs, which contributed to increases in all 3 of these hospital outcomes.

At first glance, the exponential increase of FQ could potentially be attributed to the relatively recent introduction of the ICD-9-CM code in 2008 and its increasing recognition and adoption by physicians and coders over the years. In other words, we suggest the diagnosis of FQ in the beginning of the study period was underreported, which was obvious by the low number of cases in 2008 that required us to exclude that year from trend analyses. As the definition of the FQ diagnosis suggests, the increased frailty of this population, as seen by the increasing CCI, correlates with the increasing incidence of FQ.

With a mean age of 72 years, most patients with documented FQ are covered by Medicare (82%), suggesting FQ is primarily a financial burden for Medicare. At the time of this writing, not enough data was available to depict a statistically significant increasing trend in costs. However, absolute mean costs were increasing after adjusting for inflation ($18,735 in 2009 to $20,672 in 2014, yet with overlapping standard errors).

Furthermore, in matched analyses, we exhibited an independent effect of FQ in LOS and hospital costs and charges. The matched sample did not include patients with any number of procedures (diagnostic or operative) to avoid confounding LOS, charges, and costs. The reason we excluded those patients was that matching to that extent while maintaining reasonable accuracy would yield a small matched sample and eventually compromise the power of the regression model. Nevertheless, an approximate 18% increase in costs and a 41% increase in LOS in patients without any procedures clearly shows that FQ independently leads to increased hospital resource utilization. Whether the number and type of diagnostic and operative procedures can mask or exacerbate the effect of FQ on these outcomes is a question warranting answers in future studies involving more patients.

Lastly, there is a clear association between discharge to long-term care facilities and LOS, as well as costs, among patients with FQ. These patients stay hospitalized longer and carry a higher cost burden to the hospital, even after adjusting for LOS, age, comorbidities, number of procedures, and hospital characteristics. Our exact-matched analysis revealed that FQ alone leads to longer LOS and increased costs. Therefore, earlier discharge to long-term care facilities could potentially decrease hospital resource utilization. 

Current payment models for acute care hospitals were not designed for extended stays. In contrast, long-term care facilities such as LTACHs, where enhanced nutritional support and increased physical therapy hours are available, are better equipped to care for patients with complex comorbidities and prolonged stays.²¹ Once acute conditions resolve, prolonged stays in acute care settings can expose patients with FQ to hospital-acquired conditions (eg, venous thromboembolism, decubitus ulcers, aspiration pneumonia), unnecessary testing and procedures, and even overdiagnosis. In older adults in particular, improved quality of life is important to consider after acute illness is over. Another clue indicating stagnant use of long-term-care facilities in patients with FQ is the fact that, while LTACH use has been increasing overall, our analysis revealed no increasing trend.²² Therefore, we recommend clinicians be aware of FQ and discharge patients to long-term facilities in a timely manner.

A major limitation of this study is the potential for coding inaccuracies, which are possible in any administrative database. There are still no studies that have validated the ICD-9-CM code of FQ. However, given its precise coding definition from CMS, any erroneous coding of FQ is likely due to misdiagnosis by the physician and not an error of the coder—such misdiagnoses are expected to be randomly distributed evenly among NIS discharges. NIS does not distinguish by long-term care facility type, so we could not analyze SNF and LTACH separately. We assumed SNFs and LTACHs have a similar capacity to care for patients who require prolonged hospital stay. Lastly, we recognize that there is no data describing the extent and severity of FQ (or any other diagnoses) in the NIS dataset. This makes it impossible to adjust for FQ being completely or partially resolved during the hospital stay in patients discharged to home vs long-term care facilities. However, the analysis did adjust for other comorbidities, which may correlate with the degree of FQ in these patients.

Conclusion

In conclusion, FQ is increasingly being recognized and documented and has an independent effect in increasing hospital resource utilization. Protocols that prevent or minimize the financial extent of FQ, such as facilitating earlier discharge to long-term care facilities, could potentially diminish this effect. Further studies are needed to evaluate the effect of FQ in patients using specific disease cohorts, as well as in those who undergo major operative or diagnostic procedures. 

Supplementary materials for this article are available below the References. 

References

1. 2015 ICD-9-CM Diagnosis Code 780.72—functional quadriplegia. ICD9Data.com. https://www.icd9data.com/2015/Volume1/780-799/780-789/780/780.72.htm. Accessed January 22, 2019.

2. Centers for Disease Control and Prevention (CDC). ICD-9-CM Tabular Addenda. CDC website. https://www.cdc.gov/nchs/data/icd/icdtab09add.pdf. ublished October 1, 2008. Accessed January 22, 2019.

3. Centers for Disease Control and Prevention (CDC). ICD-10-CM Official Guidelines for Coding and Reporting, FY 2017 (October 1, 2016-September 30, 2017). CDC website. https://www.cdc.gov/nchs/data/icd/10cmguidelines_2017_final.pdf. Accessed January 22, 2019.

4. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ. Long-term medical complications after traumatic spinal cord injury: a regional model systems analysis. Arch Phys Med Rehabil. 1999;80(11):1402-1410.

5. Sugarman B. Medical complications of spinal cord injury. Q J Med. 1985;54(213):3-18.

6. Healthcare Cost and Utilization Project (HCUP). Overview of the National (Nationwide) Inpatient Sample (NIS). HCUP website. https://www.hcup-us.ahrq.gov/nisoverview.jsp. Accessed January 22, 2019.

7. Healthcare Cost and Utilization Project (HCUP). NIS Database Documentation. HCUP website. https://www.hcup-us.ahrq.gov/db/nation/nis/nisdbdocumentation.jsp. Accessed January 22, 2019.

8. Patel NJ, Deshmukh A, Pant S, et al. Contemporary trends of hospitalization for atrial fibrillation in the United States, 2000 through 2010: implications for healthcare planning. Circulation. 2014;129(23):2371-2379.

9. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130-1139.

10. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173(6):676-682.

11. Healthcare Cost and Utilization Project (HCUP). Cost-to-Charge Ratio Files: User Guide for National Inpatient Sample (NIS) CCRs. HCUP website. https://www.hcup-us.ahrq.gov/db/state/CCR_NIS_UserGuide_2001-2016.pdf. Updated June 26, 2018.

12. US Bureau of Labor Statistics. CPI Inflation Calculator. Bureau of Labor Statistics website. bls.gov https://www.bls.gov/data/inflation_calculator.htm. Accessed January 27, 2019.

13. US Centers for Medicare & Medicaid Services (CMS). List of MS-DRGs, Relative Weighting Factors, & Geometric & Arithmetic Mean Length of Stay. 2008. CMS website. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Downloads/FY_08_FR_Table_5.zip.

14. US Centers for Medicare & Medicaid Services (CMS). List of MS-DRGs, Relative Weighting Factors, & Geometric & Arithmetic Mean Length of Stay. 2009. CMS website. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Downloads/FY09_IPPS_-FR_Table_5.zip.

15. US Centers for Medicare & Medicaid Services (CMS). List of MS-DRGs, Relative Weighting Factors, & Geometric & Arithmetic Mean Length of Stay. 2010. CMS website. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Downloads/FY_2010_FR_Table_5.zip.

16. US Centers for Medicare & Medicaid Services (CMS). List of MS-DRGs, Relative Weighting Factors, & Geometric & Arithmetic Mean Length of Stay. 2011. CMS wesite. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Downloads/FY_2011_FR_Table_5.zip.

17. US Centers for Medicare & Medicaid Services (CMS). List of MS-DRGs, Relative Weighting Factors, & Geometric & Arithmetic Mean Length of Stay. 2012. CMS website. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Downloads/FY_12_FR_Table_5.zip.

18. US Centers for Medicare & Medicaid Services (CMS). List of MS-DRGs, Relative Weighting Factors, & Geometric & Arithmetic Mean Length of Stay. 2013. CMS website. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Downloads/FY_13_FR_Table_5.zip.

19. US Centers for Medicare & Medicaid Services (CMS). List of MS-DRGs, Relative Weighting Factors, & Geometric & Arithmetic Mean Length of Stay. 2014. CMS website. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Downloads/FY_14_FR_Table_5.zip.

20. Blackwell M, Iacus S, King G, Porro G. CEM: coarsened exact matching in stata. Gary King website. https://gking.harvard.edu/files/gking/files/cem-stata.pdf. Published February 22, 2010. Accessed January 27, 2019.

21. Eskildsen MA. Long-term acute care: a review of the literature. J Am Geriatr Soc. 2007;55(5):775-779.

22. Medicare Payment Advisory Commission. Medicare Payment Policy: Long-Term Acute Care Hospital Services. https://www.medpac.gov/docs/default-source/reports/mar14_entirereport.pdf. Published March 14, 2014. Accessed January 2, 2019.