SEARCH

SEARCH BY CITATION

Keywords:

  • academic medical centers;
  • quality of healthcare;
  • resident education;
  • seasons;
  • spinal metastases;
  • spinal surgery;
  • surgical education;
  • surgical residency

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

BACKGROUND:

Despite widespread belief that patients admitted to teaching hospitals in July—the beginning of the academic year—have inferior outcomes, there has been little evidence to support the existence of the July phenomenon. Moreover, the impact of July admission on the outcomes after surgery for spinal metastases has not been investigated.

METHODS:

Data from the Nationwide Inpatient Sample (2005-2008) were retrospectively extracted. Patients who underwent surgery for metastatic spinal disease and were admitted to a teaching hospital were included. Multivariate logistic regression was conducted to calculate the odds of in-hospital death, the occurrence of an intraoperative complication, and the development of a postoperative complication depending on whether admission was in July or between August and June. All analyses were adjusted for differences in patient age, sex, comorbidities, primary tumor histology, visceral metastases, myelopathy, insurance status, hospital volume, and admission type.

RESULTS:

A total of 2920 admissions were evaluated. In-hospital mortality was higher in July compared with between August and June—7.5% versus 4.2%. The adjusted odds of in-hospital death were significantly higher for patients admitted in July (odds ratio [OR], 1.81; 95% confidence interval [CI], 1.13-2.91; P = .01). Patients admitted in July were significantly more likely to develop an intraoperative complication (OR, 2.11; 95% CI, 1.41-3.17; P < .001), but not a postoperative complication (OR, 1.08; 95% CI, 0.81-1.45; P = .60).

CONCLUSIONS:

In this nationwide study based on an administrative database, patients undergoing surgery for metastatic spinal disease at teaching hospitals in July had higher rates of in-hospital mortality and intraoperative complications. Cancer 2012;. © 2011 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Every July, there is an influx of new medical trainees at academic medical centers across North America as interns, residents, and fellows all simultaneously assume an increased level of responsibility at the beginning of the academic year. Many authors have expressed concern that there may be a reduced quality or decreased efficiency of care provided at teaching hospitals in July, which is referred to as the “July phenomenon” or the “July effect.”1-26 However, despite widespread speculation that there may be a July phenomenon at academic medical centers, to date there has been little evidence in support of its existence.4

The majority of studies examining medical, obstetric, trauma, critical care, and surgical patients at teaching hospitals have not found inferior outcomes at the beginning of the academic medical year.7-24, 27 Phillips and Barker analyzed national death certificates and found that there was a higher rate of medication-related errors in July,2 but it was difficult to determine whether these differences were truly related to variations in the quality of care at academic medical centers or other factors. One large study that examined patients undergoing a wide range of surgical procedures found that those admitted in July had a significantly higher mortality,5 but there may have been an insufficient adjustment of case mix and urgency of admission in their analysis. Other studies have found that admission to teaching hospitals in July is associated with increased rates of errors1-3 or complications.4, 8, 23 However, no study has specifically examined the outcomes of patients undergoing spinal surgery, including for metastatic spinal disease, in July at academic medical centers.

Patients with spinal metastases undergoing surgery may be particularly affected by variations—including seasonal and temporal differences—in quality of care. Although the treatment of spinal metastases is primarily palliative, patients with spinal cord compression, neurological dysfunction, or a pathological fracture may be treated surgically, often on an emergent or urgent basis.28-30 The extensive spinal fusion that may be performed for metastatic spinal disease is technically challenging and associated with a steep learning curve. Patients with spinal metastases may also be at high risk for the development of postoperative complications, because of widespread disease or multiple comorbidities. We report the first nationwide study examining the perioperative outcomes of patients with metastatic disease who undergo spinal surgery at teaching hospitals in July compared with those admitted between August and June.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Data Source

Data were retrospectively extracted from the Nationwide Inpatient Sample (Healthcare Cost and Utilization Project, Agency for Healthcare Research and Quality), database from 2005 to 2008. A 20% stratified sample of all nonfederal hospitals, the Nationwide Inpatient Sample is the largest all-payer national inpatient database and contains approximately 8 million annual discharges from >1000 hospitals. The stratification of the Nationwide Inpatient Sample is based on hospital characteristics: geographic region, urban or rural location, teaching status, ownership, and hospital size. The Nationwide Inpatient Sample has been used extensively to evaluate patients undergoing spinal surgery.31, 32

Inclusion Criteria

Patients were included if they had metastatic spinal disease and underwent spinal decompression with or without fusion. The International Classification of Diseases, 9th Edition, Clinical Modification (ICD-9-CM) diagnosis codes of 198.3, 198.4, and 198.5 were used to identify patients with a secondary malignant spinal tumor. The ICD-9-CM procedure codes were used to extract patients who had undergone spinal decompression (03.01, 03.09, 03.4, 03.53) or fusion (81.00-81.08, 81.61). To decrease the likelihood of misclassification of the procedure, patients were only included if 1 of the first 3 coded procedures was spinal surgery.

Separate analyses were performed for those admitted to a teaching hospital and a nonteaching hospital. The Nationwide Inpatient Sample defines a teaching hospital as 1 that meets 1 of the following 3 criteria: 1) residency training approval by the Accreditation Council for Graduate Medical Education, 2) membership in the council of teaching hospitals, or 3) a ratio of interns and residents to beds of 0.25 or higher.

Patient and Admission Characteristics

Independent variables including patient age, sex, comorbid disease, primary tumor histology, other sites of metastatic disease, myelopathy, expected primary payer, hospital volume, and admission type were extracted. Comorbidities were assessed by evaluating the effect of the categories of comorbid disease defined by Elixhauser et al33; however, paralysis, other neurological deficit, metastatic disease, lymphoma, and primary tumor were not examined because of their association with spinal metastases. Primary tumor histology was also categorized using ICD-9-CM data as 1) lung (162, V10.11); 2) breast (174, V10.3); 3) prostate (185, V10.46); 4) renal cell cancer (189.0, V10.52); and 5) other, unknown, or unspecified. Visceral (nonspinal) metastases were determined using the ICD-9-CM series of 197 and 198 (198.0, 198.1, 198.2, 198.6, 198.7, and 198.8). Data on myelopathy were assessed to indicate a greater extent of metastatic spinal disease, and was extracted using the ICD-9-CM codes 336.3, 336.8, and 336.9.

Expected primary payer is coded directly into the Nationwide Inpatient Sample, categorized as 1) Medicare, 2) Medicaid, 3) private insurance (including health maintenance organization), 4) self-pay, 5) no charge, and 6) other. Hospital volume was determined using the unique identification number for the hospital. The number of operations performed at that hospital for spinal metastases over the 4-year period was assessed; for consistency with prior reports,34 hospitals were divided based on volume into quintiles, such that each quintile had approximately 20% of the entire population. Admission type is coded as 1) elective or 2) nonelective.

Month of hospital admission is coded directly into the Nationwide Inpatient Sample. Analyses compared patients admitted in July to those admitted between August and June. For consistency with prior reports examining the July phenomenon, analyses also compared patients admitted in either July or August to those admitted between September and June.7, 8, 13, 18 In addition, analyses were performed comparing patients admitted in July to those only admitted from January to June, when new trainees may be more experienced.

Outcomes

The outcomes evaluated were in-hospital mortality, the occurrence of at least 1 intraoperative or implant complication, the development of at least 1 postoperative complication, length of hospital stay, total hospital charges, and discharge disposition. Intraoperative or implant complications were extracted using the ICD-9-CM codes for a complication related to a neurological or orthopedic implant (996.2, 996.40, 996.42, 996.47, 996.49, 996.63, 996.66, 996.67, 996.75, 996.78, 996.79) as well as accidental puncture or laceration during a procedure (998.2). Postoperative complications were extracted using ICD-9-CM codes for postprocedural neurological complications (997.00-997.09); pulmonary complications excluding pulmonary emboli (518.5, 518.81, 518.84, 997.3); venous thromboembolic events (415.11-415.19, 453.40-2, 453.8, 453.9); cardiac complications including myocardial infarction (997.1, 410); urinary and renal complications including acute renal failure (584.5, 584.9, 997.5); gastrointestinal complications (008.45, 560.1, 997.4); infectious complications including a urinary tract infection (595.0, 595.9, 599.0), meningitis (320), pneumonia (481,482,486), or a wound infection (998.32, 998.51, 998.59, 998.6, 998.81, 998.83); and decubitus ulcers (707.01-09).

Although the Nationwide Inpatient Sample does not include information on cost, data are included on total hospital charges, which exclude professional fees. The Nationwide Inpatient Sample classifies discharge disposition into 1) routine (to home), 2) transfer to a short-term hospital, 3) other transfer, 4) home health care, 5) against medical advice, 6) died, and 7) unknown. A nonroutine discharge was defined any disposition other than the first category.

Statistical Analysis

Descriptive statistics were conducted for demographic and hospital variables. Categorical and continuous variables were compared using chi-square and 2-tailed Student t test, respectively. Multivariate logistic regression models were constructed for binomial variables and multivariate linear regression was performed to analyze length of hospital stay and total hospital charges. Because of the positive skew of the continuous variables, logarithmic transformation was performed. All multivariate analyses were adjusted for the patient characteristics. Statistical analyses were performed using Stata version 11 (StataCorp, College Station, TX). All analyses were performed accounting for the complex sampling technique of the Nationwide Inpatient Sample (using svy commands), which incorporates clustering at the hospital level and the weight of each discharge. A P value of <.05 was accepted as significant.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Demographics of the Study Population

A total of 2920 admissions at teaching hospitals were included, of which 9.6% occurred in July. The demographics of the patients admitted to teaching hospitals in July are compared with those admitted during the remainder of the year in Table 1. There were no significant differences in the majority of the patient characteristics, with the exception of expected primary payer (Table 1). A total of 968 admissions at nonteaching hospitals were also evaluated, of which 8.0% were in July. The demographics of patients admitted to a nonteaching hospital in July were not significantly different from those admitted between August and June (Table 2).

Table 1. Demographics of the Patients Undergoing Spinal Surgery for Metastatic Disease at Teaching Hospitals Admitted in July Compared With Those Admitted Between August and June
CharacteristicTotal Population, n=2920July Admission, n=279August-June Admission, n=2641P
  • Abbreviation: SD, standard deviation.

  • All data are presented as percentages, with the exception of age.

  • a

    Statistically significant difference.

Age, mean y (SD)58.3 (14.8)59.6 (14.2)58.2 (14.8).12
Sex   .24
 Women45.642.346.0 
 Men54.457.754.0 
Comorbid disease   .56
 0 comorbid diseases27.730.827.3 
 1 comorbid disease28.028.028.0 
 2 comorbid diseases22.822.222.9 
 3+ comorbid diseases21.519.021.8 
Myelopathy33.935.533.7.56
Visceral metastases30.128.730.3.58
Primary tumor histology   .06
 Lung21.124.720.8 
 Breast15.412.215.5 
 Prostate10.614.010.2 
 Renal cell11.38.611.6 
 Other or missing41.940.542.0 
Expected primary payer   .002a
 Medicare35.840.935.2 
 Medicaid11.211.811.1 
 Private insurance46.940.547.6 
 Self-pay2.53.22.5 
 No charge0.31.40.2 
 Other3.32.23.4 
Hospital volume   .95
 Very low21.219.721.4 
 Low20.921.220.9 
 Medium18.620.118.4 
 High20.720.420.7 
 Very high18.618.618.6 
Admission type   .97
 Nonelective63.463.463.3 
 Elective36.736.636.7 
Table 2. Demographics of the Patients Undergoing Spinal Surgery for Metastatic Disease at Nonteaching Hospitals Admitted in July Compared With Those Admitted Between August and June
CharacteristicTotal Population, n=968July Admission, n=77August-June Admission, n=891P
  1. Abbreviation: SD, standard deviation.

  2. All data are presented as percentages, with the exception of age.

Age, mean y (SD)62.0 (13.2)59.6 (16.0)62.2 (12.9).10
Sex   .29
 Men55.261.054.7 
 Women44.839.045.3 
Comorbid disease   .46
 0 comorbid diseases22.423.422.3 
 1 comorbid disease27.832.527.4 
 2 comorbid diseases25.727.325.6 
 3+ comorbid diseases24.116.924.7 
Myelopathy30.129.930.1.97
Visceral metastases26.924.727.1.65
Primary tumor histology   .40
 Lung21.822.121.8 
 Breast17.715.617.9 
 Prostate14.410.414.7 
 Renal cell9.96.510.2 
 Other or missing36.345.535.5 
Expected primary payer   .65
 Medicare45.046.844.9 
 Medicaid6.97.86.9 
 Private insurance43.444.243.3 
 Self-pay2.40.02.6 
 No charge0.00.00.0 
 Other2.31.32.4 
Hospital volume   .86
 Very low26.229.925.9 
 Low14.611.714.8 
 Medium16.515.616.6 
 High16.314.316.5 
 Very high26.328.626.2 
Admission type   .77
 Nonelective61.359.761.5 
 Elective38.740.338.5 

In-hospital Mortality

A total of 4.4% of patients died postoperatively at teaching hospitals during the hospitalization. The seasonal variation of in-hospital mortality at academic medical centers is shown in Figure 1 (Top). The crude mortality was higher for patients admitted to teaching hospitals in July when compared with the remainder of the year—7.5% versus 4.1%. The adjusted odds of in-hospital death were significantly higher for patients admitted to teaching hospitals in July (odds ratio [OR], 1.81; 95% confidence interval [CI], 1.13-2.91; P = .01, Table 3). The crude mortality was also higher for patients admitted to academic medical centers in July or August when compared with the remainder of the year—6.6% versus 4.0%. The adjusted odds of in-hospital death were significantly higher for patients admitted to academic medical centers during either July or August (OR, 1.76; 95% CI, 1.18-2.62; P = .006) compared with those admitted between September and June. In addition, the crude mortality was higher for patients admitted to teaching hospitals in July (7.5%) when compared with those admitted between January and June (3.7%). The adjusted odds of in-hospital death were significantly higher for those admitted in July compared with those admitted between January and July (OR, 2.00; 95% CI, 1.19-3.36; P = .009). Moreover, a multivariate analysis testing the interaction between July admission and the performance of surgery at a teaching hospital for the entire study population (those admitted to a teaching and to a nonteaching hospital) also showed significantly higher in-hospital mortality for those admitted to academic medical centers in July (OR, 1.80; 95% CI, 1.12-2.90; P = .02).

thumbnail image

Figure 1. Depiction is shown of the seasonal variation in the crude rates (and the associated standard error) of mortality after surgery for spinal metastases at teaching (Top) and nonteaching (Bottom) hospitals. Given that the academic year at academic medical centers begins in July, the seasons were divided into July to September (summer), October to December (fall), January to March (winter), and April to June (spring).

Download figure to PowerPoint

Table 3. Comparison of the Outcomes After Surgery for Spinal Metastases at Teaching Hospitals in the United States Depending on Whether Admission Was in July or Between August and June
OutcomeJuly Admission, n=279August-June Admission, n=2641Adjusted Odds Ratioa95% Confidence IntervalP
  • a

    All analyses are adjusted for differences in patient age, sex, and comorbid disease, the proportion of patients with myelopathy, primary tumor histology, visceral metastases, expected primary payer, hospital volume of surgery for metastatic spine disease (by quintiles), and admission type (elective vs nonelective).

  • b

    Statistically significant difference.

In-hospital mortality, %7.54.21.811.13-2.91b.01b
Intraoperative or implant complication rate, %8.24.02.111.41-3.17b<.001b
Postoperative complication rate, %35.533.21.080.81-1.45.60
Nonroutine hospital discharge, %67.761.41.310.96-1.80.09

In-hospital mortality was 4.7% at nonteaching hospitals; the seasonal variation in mortality at nonteaching hospitals is shown in Figure 1 (Bottom). Mortality was lower at nonteaching hospitals in July (1.3%) compared with between August and June (5.0%). The adjusted odds of in-hospital death were significantly lower for patients admitted to a nonteaching hospital in July (OR, 0.16; 95% CI, 0.03-0.75; P = .02; Table 4). Mortality was also lower during July and August when compared with the remainder of the year—3.3% versus 4.9%. The adjusted odds of in-hospital mortality were not significantly different for patients admitted to nonteaching hospitals in July or August (OR, 0.95; 95% CI, 0.35-2.55; P = .91). Moreover, the crude mortality for patients admitted to nonteaching hospitals was lower in July (1.3%) when compared only to those admitted between January and June (4.3%). The adjusted odds of in-hospital death were significantly lower for patients admitted to nonteaching hospitals in July compared with those admitted between January and June (OR, 0.18; 95% CI, 0.04-0.89; P = .04).

Table 4. Comparison of the Outcomes After Surgery for Spinal Metastases at Nonteaching Hospitals in the United States Depending on Whether Admission Was in July or Between August and June
OutcomeJuly Admission, n=77August-June Admission, n=891Adjusted Odds Ratioa95% Confidence IntervalP
  • a

    All analyses are adjusted for differences in patient age, sex, and comorbid disease, the proportion of patients with myelopathy, primary tumor histology, visceral metastases, expected primary payer, hospital volume of surgery for metastatic spine disease (by quintiles), and admission type (elective vs nonelective).

  • b

    Statistically significant difference.

In-hospital mortality, %1.35.00.160.03-0.75b.02b
Intraoperative or implant complication rate, %3.93.31.040.25-4.32.96
Postoperative complication rate, %23.433.30.580.33-1.00.05
Nonroutine hospital discharge, %64.969.31.000.57-1.76.99

Intraoperative Complications

Intraoperative or implant complications occurred in 4.4% of patients admitted to academic medical centers. A higher proportion of patients admitted in July to teaching hospitals had an intraoperative or implant complication compared with those admitted during the remainder of the year—8.2% versus 4.0%, respectively (Fig. 2). The adjusted odds of the development of an intraoperative or implant complication were significantly higher for those admitted in July (OR, 2.11; 95% CI, 1.41-3.17; P < .001). Likewise, a higher proportion of patients admitted in July or August to teaching hospitals had an intraoperative complication (5.8%) compared with patients admitted between September and June (3.5%). The adjusted odds of the occurrence of an intraoperative complication were significantly higher for patients admitted in July or August (OR, 1.80; 95% CI, 1.20-2.70; P = .004). Moreover, the crude rates of an intraoperative complication at academic medical centers were higher for patients admitted in July (8.2%) compared with those admitted between January and June (4.4%). The adjusted odds of an intraoperative complication at teaching hospitals were significantly higher for patients admitted in July compared with those admitted between January and June (OR, 1.91; 95% CI, 1.21-3.00; P = .005). In addition, a multivariate analysis testing the interaction of July admission and the performance of surgery at a teaching hospital for the entire study population showed that the development of an intraoperative complication was significantly more common for those admitted in July to academic medical centers (OR, 2.15; 95% CI, 1.42-3.28; P < .001).

thumbnail image

Figure 2. Variations in the crude percentage (and the associated standard error) of patients who experienced an intraoperative or implant complication at teaching (Top) and nonteaching (Bottom) hospitals are shown. Given that the academic year at academic medical centers begins in July, the seasons were divided into July to September (summer), October to December (fall), January to March (winter), and April to June (spring).

Download figure to PowerPoint

Intraoperative or implant complications were more common in patients admitted to nonteaching hospitals in July—3.9% versus 3.3%. The adjusted odds of an intraoperative complication did not differ significantly at nonteaching hospitals for patients admitted in July compared with those admitted between August and June (OR, 1.08; 95% CI, 0.27-4.41; P = .91). A total of 3.3% of patients admitted to nonteaching hospitals in July or August and 3.2% of patients admitted between September and June had an intraoperative or implant complication. Likewise, the adjusted odds of an intraoperative or implant complication did not differ significantly for those admitted to nonteaching hospitals between July or August (OR, 0.95; 95% CI, 0.35-2.60; P = .92). In addition, the crude rate of an intraoperative complication was slightly higher for patients admitted to a nonteaching hospital in July (3.9%) compared with between January and June (3.5%). The adjusted odds of the development of an intraoperative or implant complication were not significantly different for those admitted to a nonteaching hospital in July compared with those admitted between January and June (OR, 1.13; 95% CI, 0.23-5.60; P = .89).

Other Outcomes

Postoperative complications occurred in 32.7% of patients admitted to teaching hospitals and in 31.7% of patients admitted to a nonteaching hospital. The adjusted odds of the development of at least 1 postoperative complication were not significantly different for patients who were admitted in July to teaching hospitals (Table 3) or to nonteaching hospitals (Table 4). The median length of hospital stay was 9 (interquartile range, 6-15) days at teaching hospitals and 8 (interquartile range, 5-14) days at nonteaching hospitals. Length of hospital stay did not differ significantly for patients admitted in July to teaching (Table 5) or nonteaching hospitals (Table 6). The median total hospital charge was $97,287 (interquartile range, $55,082-$165,162) at teaching hospitals and $83,359 (interquartile range, $46,104-$141,188) at nonteaching hospitals. Total hospital charges did not differ significantly based on the month of admission for those at teaching hospitals (Table 5) or at nonteaching hospitals (Table 6). A total of 62.0% of patients admitted to a teaching hospital and 70.0% of patients admitted to a nonteaching hospital had a nonroutine hospital discharge. The adjusted odds of a nonroutine discharge were not significantly different for those admitted in July either to teaching hospitals (Table 3) or to nonteaching hospitals (Table 4).

Table 5. Comparison of the Efficiency of Care for Patients With Spinal Metastases Who Underwent Surgery at an Academic Medical Center Depending on Whether the Month of Admission Was July or Between August and June
OutcomeJuly Admission, n=279August-June Admission, n=2641Adjusted Percentage Differencea95% Confidence IntervalP
  • Abbreviation: IQR, interquartile range.

  • a

    All analyses are adjusted for differences in patient age, sex, and comorbid disease, the proportion of patients with myelopathy, primary tumor histology, visceral metastases, expected primary payer, hospital volume of surgery for metastatic spine disease (by quintiles), and admission type (elective vs nonelective).

Length of hospital stay, median d (IQR)9 (6-16)9 (6-15)2.22%−1.65, 6.08.26
Total hospital charges, median $ (IQR)100,931 (58,487-183,158)96,780 (54,793-164,045)1.96%−2.82, 6.73.42
Table 6. Comparison of the Efficiency of Care for Patients With Spinal Metastases Who Underwent Surgery at Nonteaching Hospitals Depending on Whether the Month of Admission was July or Between August and June
OutcomeJuly Admission, n=77August-June Admission, n=891Adjusted Percentage Differencea95% Confidence IntervalP
  • Abbreviation: IQR, interquartile range.

  • a

    All analyses are adjusted for differences in patient age, sex, and comorbid disease, the proportion of patients with myelopathy, primary tumor histology, visceral metastases, expected primary payer, hospital volume of surgery for metastatic spine disease (by quintiles), and admission type (elective vs nonelective).

Length of hospital stay, median d (IQR)7 (4-15)8 (5-14)−2.88%−10.22, 4.46.44
Total hospital charges, median $ (IQR)64,925 (34,339-143,509)83,461 (47,327-142,625)−4.75%−12.85, 3.35.25

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Spinal metastases are an important cause of morbidity and mortality in oncology patients; an estimated 5% to 10% of patients with cancer will develop metastatic epidural spinal cord compression.28 Although pain is the most common presenting symptom of patients with spinal metastases, those with metastatic epidural spinal cord compression may also present with radiculopathy or myelopathy. Spinal metastases may cause a severely impaired quality of life, and patients with metastatic epidural spinal cord compression are also at risk for vertebral collapse, spinal instability, and further compression of the spinal cord.28-30 Radiation therapy, percutaneous spinal procedures (vertebroplasty or kyphoplasty), stereotactic radiosurgery, surgical decompression with or without fusion, or a combination of these modalities may be appropriate depending on the primary tumor type, prognosis, and spinal stability.35 Indications for surgery include rapid neurological deterioration; neural compression because of spinal deformity, osseous, or tumor pathology; radioresistant tumors; and recurrence after maximal radiotherapy.30, 35 The decision to pursue surgery rather than another treatment modality should not vary based on the month of hospital admission.

In this study, 2920 patients from across the United States who underwent surgery for metastatic spinal disease between 2005 and 2008 were examined to evaluate the impact of July hospital admission—the beginning of the academic year—at teaching hospitals. Most of the patient demographics, including age, sex, comorbidities, primary tumor histology, other metastases, and admission type, did not differ significantly when comparing patients admitted in July to those admitted between August and June. However, those admitted in July were less likely to have private insurance.

Patients admitted to academic medical centers during July had significantly higher adjusted odds of in-hospital death, as well as of developing an intraoperative or implant complication, compared with those admitted between August and June. Similar analyses found that patients admitted in either July or August also had higher adjusted odds of in-hospital mortality and the development of an intraoperative or implant complication compared with those admitted between September and June. In contrast, the likelihood of both in-hospital death and the occurrence of an intraoperative or implant complication were not significantly higher at nonteaching hospitals during July when compared with August to June or during July and August when compared with September to June.

The reasons for the differential outcomes of patients who present to academic medical centers in July are likely complex and multifactorial; the data available in the Nationwide Inpatient Sample do not allow for a detailed determination of which factors are primarily responsible for the July phenomenon. The relative inexperience of new interns, residents, and fellows at the beginning of the academic year may be a factor.1, 3 Although senior residents and staff physicians may more vigilantly supervise trainees during this transition period,9 this may be insufficient to prevent some complications. Working in a new hospital environment, where there may be differences in electronic medical records, ordering (and accessing) diagnostic tests, and interactions with other health professionals, may be another factor that is independent of experience.3, 4 Moreover, other health professionals, including nurses, operating room staff, respiratory therapists, and pharmacists, may be beginning their career or working at a new institution during the summer.7 Finally, key senior personnel, including staff physicians and nursing staff, may schedule vacation during the summer, potentially further decreasing supervision.4, 7

Although differences in the quality of care at teaching hospitals at the beginning of the academic year may contribute to the July phenomenon, patients admitted in July may also have been sicker (such as having more widespread metastases) than patients admitted during the remainder of the year. However, there were no significant differences in the number of comorbid diseases, admission type, primary tumor histology, or the percentage of patients with nonspinal metastases between those admitted to teaching hospitals in July compared with those admitted between August and June. In addition, surgeons at nonteaching hospitals may be more likely to be on vacation in July, necessitating that patients with emergent or urgent presentations be treated at teaching hospitals during this time period. This may partially explain the increased mortality at teaching hospitals and the decreased mortality at nonteaching hospitals in the summer.

There are many advantages of the usage of the Nationwide Inpatient Sample to investigate the July phenomenon, and this database has been used to evaluate the impact of July hospital admission for patients with other conditions.7, 13, 15, 16 The Nationwide Inpatient Sample provides a nationwide perspective by sampling a broad spectrum of hospitals across the country. The inclusion of both academic medical centers and nonteaching hospitals allows for the impact of July hospital admission at both types of institutions to be evaluated. Data on several confounding variables—including comorbid diseases, primary tumor histology, other (nonspinal) metastases, myelopathy, and socioeconomic factors—can be extracted and adjusted for in multivariate regression analyses.

There are also some important limitations of this study. The Nationwide Inpatient Sample does not include information on the type of training programs at an institution; thus, it was not possible to restrict the study to hospitals with either an orthopedic or a neurosurgical residency or fellowship program. However, even if a hospital does not have an orthopedic or neurosurgical training program, patients may interact with resident physicians in the emergency room, in the operating room (if there is an anesthesia residency program), and the intensive care unit—all of which may impact outcomes at the beginning of the academic medical year.7 Moreover, the Nationwide Inpatient Sample only has limited information on patients at presentation; thus, it was not possible to adjust for the specific chief complaint, neurological examination, imaging findings, or expected prognosis. The database only includes information on patients during the hospitalization. Therefore, the impact of July hospital admission on long-term neurological function and mortality could not be assessed. As with all administrative databases, there may be miscoded data, but there is no reason to suspect that miscoding would preferentially affect those admitted during a given month.

Future studies—especially with prospectively collected data—are needed not only to confirm the existence of a July phenomenon in patients presenting with other conditions, but also to investigate potential interventions to partially alleviate the July phenomenon. Nonetheless, this nationwide study suggests that patients undergoing surgery for metastatic spinal disease may have inferior outcomes at the beginning of the academic year at academic medical centers.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

H.H.D. is supported by the Doris Duke Clinical Research Fellowship for Medical Students at Johns Hopkins University. There are no other relevant sources of funding for this article.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Z.L.G. has stock ownership in US Spine and Spinal Kinetics; he has research support from AO Spine North America, Integra Life Sciences, and Johnson & Johnson (DePuy Spine); he also has received fellowship support, a stipend as a board member, and an honorarium from AO Spine North America. A.B. is the recipient of a grant from Johnson & Johnson (DePuy Spine).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES