- Email: [email protected]
Retained Foreign Bodies: Risk and Outcomes at the National Level
Retained Foreign Bodies: Risk and Outcomes at the National Level Zaid H Al-Qurayshi, MBChB, MPH, Adam T Hauch, Emad Kandil, MD, FACS
MD, MBA,
Douglas P Slakey,
MD, MPH, FACS,
Retained foreign bodies (RFB) after operative interventions are linked to an increased risk of morbidity and mortality, and represent a medico-legal liability. We aimed to identify associated risk factors and outcomes related to iatrogenic RFB in the United States. STUDY DESIGN: A cross-sectional analysis was performed on all interventions that resulted in a secondary diagnosis of RFB in the Nationwide Inpatient Sample (NIS) from 2003 to 2009. Comparative controls were randomly selected from patients who underwent similar procedures. RESULTS: We identified 3,045 cases of RFB, and 12,592 controls were included. The majority of incidents, 968 (31.8%), were reported after gastrointestinal interventions. Risk of RFB was higher in teaching hospitals (odds ratio [OR] 1.31, 95% CI [1.19, 1.45], p < 0.001). For abdominopelvic procedures, patients admitted with traumatic injuries did not demonstrate a higher risk of RFB compared with electively admitted patients (OR 1.70, 95% CI [0.94, 3.07], p ¼ 0.08). However, for procedures unrelated to abdominopelvic surgery, patients admitted for trauma had a lower risk (OR 0.62, 95% CI [0.50, 0.78], p < 0.001). Obesity (BMI 30 kg/m2) and older age (65 years) were significantly associated with a higher risk only for abdominopelvic procedures (p < 0.01 for both). Retained foreign bodies were associated with a higher average cost of health services ($26,678.00 $769.69 vs $12,648.00 $192.80, p < 0.001). CONCLUSIONS: Retained foreign bodies have unfavorable and nationally tangible clinical and economic outcomes. The risk profile for RFB at the national level seems to demonstrate an association with demographic and clinical factors including nature of the procedure, type of admission, and trauma status. Teaching hospitals are associated with a higher risk. Targeted efforts toward identified high-risk populations are needed to avoid these morbid and costly complications. (J Am Coll Surg 2015;220:749e759. 2015 by the American College of Surgeons)
BACKGROUND:
inpatient operations in the United States.2-4 Sponges are the most commonly retained items compared with needles and instruments.3,5-7 Regardless of the nature of the foreign body, the consequences of left-behind objects include grievous health implications to patients, indefensible litigations, and irreparable damage to the reputations of health care professionals and institutions.5,6,8-10 Despite the theoretically preventable nature of RFB, tackling this medical error represents a true challenge to health care policy makers.6 Several standardized protocols and regulations have been devised to mitigate this issue, like manual and automated sponge and instrument counting, radiographic screening, and goals for improvement of communication among the surgical team; however, no method has proven to be completely successful in preventing RFB.2-7,10-19 Moreover, it is hypothesized that the
Retained foreign bodies (RFB) after interventional procedures are recognized by the National Quality Forum as preventable medical errors, listed as “never events” in their updated 2011 report Serious Reportable Events In Healthcare.1 The true incidence of RFB is still elusive; it is estimated that it ranges from 1 in every 1,000 to 1,500 abdominal operations and 1 in every 8,000 to 18,000 Disclosure Information: Nothing to disclose. Presented at the Southern Surgical Association 126th Annual Meeting, Palm Beach, FL, November 30eDecember 3, 2014. Received December 7, 2014; Accepted December 9, 2014. From the Department of Surgery, Tulane University School of Medicine, New Orleans, LA. Correspondence address: Emad Kandil, MD, FACS, Department of Surgery, Tulane University School of Medicine, 1430 Tulane Ave, Room 8510 (Box SL-22), New Orleans, LA 70112. email: [email protected]
ª 2015 by the American College of Surgeons Published by Elsevier Inc.
749
http://dx.doi.org/10.1016/j.jamcollsurg.2014.12.015 ISSN 1072-7515/14
750
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
Abbreviations and Acronyms
BMI ¼ body mass index HCUP ¼ Healthcare Cost and Utilization Project ICD-9 ¼ International Classification of Disease, 9th Revision MORPI ¼ major operating room procedure indicator NIS ¼ Nationwide Inpatient Sample OR ¼ odds ratio RFB ¼ retained foreign bodies
reason behind the continuing occurrence of this complication is that some risk factors remain unexplored.20 Studies that have addressed RFB consist predominately of case reports and case series.6,21-37 The main hindrance in conducting larger scale studies is the lack of considerable aggregate data, due in part to the natural history of RFB in which an unknown number of events can go unnoticed.8,17 Additionally, some investigators have proposed that hospitals are inclined toward under-reporting such incidents because of their critical nature.5,6,8,9,38 This becomes obvious when reviewing the literature because most of the larger analytic studies depend on data from malpractice lawsuits.2,6 To date, 3 relatively large casecontrol studies, with RFB sample sizes of less than 60, have examined the risk factors for RFB after a variety of surgical interventions.2,20,38 Among these studies, discrepancies exist in identifying persistent risk factors.2,20,38 However, in a recent meta-analysis that included the previously mentioned studies, several factors exhibited significant risk for unintentional RFB, including intraoperative blood loss of more than 500 mL, longer duration of operation, involvement of more than 1 subprocedure or surgical team, unexpected intraoperative factors, and finally, lack of or incorrect surgical count.39 In this study, we aimed to assess the frequency of RFB at the national level and across all disciplines, as opposed to previous studies that have focused on certain institutions’ experiences and specific fields. We also aimed to identify the risk factors for and the clinical and economic outcomes associated with RFB in the United States by using data from the Nationwide Inpatient Sample (NIS) for the period of 2003 to 2009.
METHODS This study is a cross-sectional analysis using the Nationwide Inpatient Sample (NIS) database for the years 2003 to 2009. The NIS is part of the Healthcare Cost and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality (AHRQ). This is the largest all-payer inpatient care database that
J Am Coll Surg
is publicly available in the United States. It contains data from approximately 8 million hospital stays from about 1,000 hospitals sampled to approximate a 20% stratified sample of US community hospitals.40 Although NIS data are available for the period between 1988 to 2010 up to the time of the data analysis, the data elements and the sampling methodologies were evolving and changing over the years, which put a technical restraint on longitudinal analyses, and for which the HCUP is prompting investigators to be cautious with data before and after the period included in this analysis.40 For these coding precautions, we opted to include the data analysis between 2003 and 2009 to perform a thorough, accurate, and unjeopardized investigation. The Clinical Classifications Software (CCS) was used to identify the primary procedures of interest41; the International Classification of Disease, 9th Revision (ICD-9) was used in defining the other parameters of the study. This software is a diagnosis and procedure categorization scheme developed by HCUP and based on the ICD-9; it provides a smaller number of clinically meaningful categories that are more useful for presenting descriptive statistics than are individual ICD-9 codes.41 The database was surveyed for patients who developed a secondary diagnosis of RFB (ICD-9: 998.5) during their hospital stay. Then, based on their primary procedure, controls were randomly selected from those who underwent the same procedures performed within the same year. The main risk factors for RFB that were assessed included: Age less than 36, 36 to 65, and greater than 65 years old; sex; site of procedure, determined based on the involved system by the primary procedure: gastrointestinal tract/spleen, cardiovascular system, respiratory system, nervous system, head and neck, eye, musculoskeletal system, urinary system, breast, female reproductive system, and male reproductive system; NIS major operating room procedure indicator (MORPI): not reported vs reported42; admission/trauma status: elective/nontrauma, nonelective/nontrauma, elective/trauma, and nonelective/trauma; body mass index (BMI) 30 kg/ m2; and hospital teaching status: nonteaching vs teaching. However, in assessing the above risk factors, more than 50% of the records were missing response values for race and MORPI. In order not to lose the power of the study, 2 multivariate logistic regression models were performed, primary model “A” excluding race and MORPI, and a secondary model “B” including race and MORPI. The results of both models are provided. Hospital characteristics that were examined for the risk of more than 1 incidence of RFB per year were: Hospital region: Northeast, Midwest, West, and South; hospital
Al-Qurayshi et al
Vol. 220, No. 4, April 2015
Table 1.
Retained Foreign Bodies: Risk and Outcomes
751
Descriptive Statistics of the Study Population in Relation to Outcomes Related to Retained Foreign Body
Variable
Age, y <36 36e65 >65 Sex Male Female Race White Black Hispanic Asian/Pacific Islander Native American Other Annual household income, $ <39,000 39,000e47,999 48,000e62,999 >62,999 Service payer Medicare Medicaid Private/HMO Self-pay No charges Other Procedure site/system GI tract/spleen Cardiovascular Respiratory Nervous Head and neck Eye Musculoskeletal Urinary Breast Female reproductive Male reproductive MORPI Not reported Reported Procedure region Non-abdominopelvic Abdominopelvic Admission/trauma status Elective/nontrauma Nonelective/nontrauma
% Study population (n ¼ 15,637)
Retained foreign body status % Cases (n ¼ 3,045) % Controls (n ¼ 12,592)
p Value*
30.6 38.2 31.2
19.3 47.7 33.0
33.3 35.9 30.8
<0.001
36.4 63.6
41.9 58.1
35.1 64.9
<0.001
69.2 11.7 12.6 2.5 0.6 3.4
72.4 11.8 10.3 2.0 0.7 2.8
68.4 11.7 13.1 2.7 0.6 3.5
<0.001
26.6 26.2 24.3 22.9
26.7 25.9 25.3 22.0
26.5 26.3 24.1 23.1
0.45
34.2 16.9 40.8 4.1 0.4 3.7
37.3 12.5 41.9 3.8 0.5 4.0
33.4 18.0 40.5 4.2 0.4 3.6
<0.001
24.2 15.5 8.0 3.0 0.8 0.1 16.1 2.9 0.5 27.6 1.2
31.8 18.3 4.9 3.4 0.6 0.1 17.1 4.6 0.3 17.6 1.3
22.3 14.8 8.8 3.0 0.8 0.1 15.9 2.5 0.5 30.1 1.2
<0.001
30.8 69.2
8.7 91.3
35.8 64.2
<0.001
51.2 48.8
50.6 49.4
51.4 48.6
0.44
44.7 49.5
51.7 43.4
43.0 51.0 (Continued)
752
Table 1.
Al-Qurayshi et al
J Am Coll Surg
Retained Foreign Bodies: Risk and Outcomes
Continued
Variable
Elective/trauma Nonelective/trauma CCIS Low: 0e1 Medium-low: 2e3 Intermediate: 4e5 High: 6 Obesity (BMI 30 kg/m2) No Yes In-hospital mortality No Yes Hospital region Northeast Midwest South West Hospital size (no. of beds) Small Medium Large Hospital location Rural Urban Hospital teaching status Nonteaching Teaching
% Study population (n ¼ 15,637)
Retained foreign body status % Cases (n ¼ 3,045) % Controls (n ¼ 12,592)
p Value*
0.6 5.2
0.5 4.5
0.7 5.3
<0.001
96.3 3.7 0.1 0.0
95.8 4.2 0.0 0.0
96.4 3.6 0.1 0.0
0.18
93.0 7.0
90.9 9.1
93.5 6.5
<0.001
97.1 2.9
96.6 3.4
97.3 2.7
0.05
19.5 23.7 36.2 20.6
19.9 23.1 35.0 22.1
19.5 23.8 36.5 20.3
0.29
10.9 24.2 64.9
11.0 23.4 65.6
10.9 24.4 64.7
0.60
10.0 90.0
10.3 89.7
9.9 90.1
0.55
49.2 50.8
43.6 56.4
50.5 49.5
<0.001
*Chi-square test. BMI, body mass index; CCIS, Charlson Comorbidity Index Score; GI, gastrointestinal; MORPI, major operating room procedure indicator.
size (number of beds): small, medium, and large42; hospital location: urban vs rural; and hospital teaching status: teaching hospital vs nonteaching. The outcomes of RFB were assessed by examining 2 factors: Absence or presence of 1 or more postoperative complications, classified into cardiovascular complications, pulmonary complications, acute renal failure, bleeding, infection/sepsis, wound complications, shock, and cystitis; and average cost of the health services on the hospital per patient. Other secondary independent factors that were assessed as potential confounders were patient race (white, black, Hispanic, Asian/Pacific Islander, Native American, and other), a modification of the Charlson Comorbidity Index Score (CCIS) (none, mild, moderate, severe),43 inpatient death, annual household income (quartile classification: <$39,000; $39,000 to $47,999;
$48,000 to $62,999; and >$62,999), main payer of health service (Medicare, Medicaid, private insurance, self-pay, no charges, and other). Statistical analysis used weighted data reflecting a national estimate. The records’ weights are available in the NIS data and calculated based on the stratification variables that were used in sampling methodology. These variables include hospital geographic region, urban or rural location, teaching status, hospital size (number of beds), and ownership. Cross-tabulation and chi-square tests were used to examine the association between each of the independent factors and the outcome of interest. Factors with significant association were considered possible confounders and were included in multivariate logistic regression models. Multivariate logistic regression models were used to calculate the odds ratio (OR) and 95% CI.
Vol. 220, No. 4, April 2015
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
753
Figure 1. Frequency of retained foreign bodies according to the type of procedure, 2003 to 2009.
Student’s t-test was used to test for the differences in cost. Stratification by the site of the procedure (abdominopelvic procedures vs non-abdominopelvic procedures) and hospital teaching status was conducted in the assessment of the factors of interest. Significance level was set as (a ¼ 0.05). All data analyses were performed using SAS 9.2 for Windows (SAS Institute Inc).
RESULTS There were 3,045 cases of RFB identified, and 12,592 controls were included (Table 1). A control:case ratio of at least 4:1 for each procedure code was included. The mean age of the study population was 50.9 years (0.3 years [SEM]), with the majority being white (69.2%), female (63.6%), and with private health insurance (40.8%) (Table 1). There were 102 RFB patients who died during their hospital stay, compared with 364 controls (p > 0.05). The most common procedure categories of RFB were for gastrointestinal procedures (968 cases; 31.8%), while interventions for the cardiovascular system, female reproductive organs, and musculoskeletal system followed with nearly identical frequencies of 554, 538, and 523 cases, respectively (Fig. 1). Overall, younger patients (less than 36 years) had a lower risk of RFB compared with older age groups (p < 0.001) (Table 2). On stratification by procedure type (Table 3), the same relationship was observed for abdominopelvic operations. In non-abdominopelvic operations, the younger age group had a marginally higher risk of RFB compared with patients older than 65 years (20% vs 17.2%, p ¼ 0.044). Males were more likely to have retained objects only for procedures in the abdominopelvic region (OR 1.18, 95% CI [1.02, 1.37], p ¼ 0.025) (Table 3). Similarly, obesity was a significant risk factor for abdominopelvic operations (OR 1.45, 95% CI [1.15, 1.83], p ¼ 0.002) (Tables 2 and 3).
Abdominopelvic procedures were significantly associated with a higher risk of RFB compared with other sites (p < 0.05 for all) (Table 2). However, there was no significant difference in the prevalence of RFB between gastrointestinal and urinary system procedures (OR 1.19, 95%CI [0.95, 1.49], p ¼ 0.13). On the other hand, patients who underwent major operating room procedures (MORPI) were at least 4 times more likely to have a retained object, regardless of the surgery site (OR 5.56, 95% CI [4.07, 7.59], p < 0.001). In general, and for interventions in the nonabdominopelvic region specifically, elective admissions and nontraumatic presentations were associated with a higher prevalence of RFB compared with nonelective and traumatic presentations (p < 0.05) (Table 2). However, for patients who required abdominopelvic procedures, there was no significant difference among patients sustaining trauma, regardless of their admission type, compared with those with elective or nontrauma status (p > 0.05). With respect to hospital characteristics, teaching hospitals had a higher incidence of RFB compared with nonteaching facilities (OR 1.31, 95% CI [1.19, 1.45], p < 0.001) (Tables 1 and 2). Furthermore, the risk of more than 1 incident of RFB per year was significantly higher in teaching and large hospitals (p < 0.05) (Table 4). The lowest incidence of recurrent RFB was in hospitals located in the Midwest, while the remaining regions (Northeast, South, and West), all had significantly higher incidences compared with the Midwest (p < 0.05) (Table 4). Cases of unintentionally retained objects, compared with controls, were twice as likely to have 1 or more postoperative complications (OR 2.49, 95% CI [2.25, 2.75], p < 0.001) (Table 5). The average cost (SEM) of hospitalization for a patient having an
754
Table 2.
Al-Qurayshi et al
J Am Coll Surg
Retained Foreign Bodies: Risk and Outcomes
Weighted Adjusted Odds Ratio of a Retained Foreign Body after a Procedure for Selected Risk Factors
Factor
Age, y <36 36e65 >65 Sex Male Female Procedure site/system GI tract/spleen Cardiovascular Respiratory Nervous Head and neck Eye Musculoskeletal Urinary Breast Female reproductive Male reproductive MORPI Not reported Reported Admission/trauma status Elective/nontrauma Nonelective/nontrauma Elective/trauma Nonelective/trauma Obesity (BMI 30 kg/m2) No Yes Hospital teaching status Nonteaching Teaching
% RFB
aOR
12.3 24.3 20.6
Reference 1.724 1.416
22.4 17.8
1.042 Reference
25.6 22.9 11.9 21.8 15.4 10.4 20.6 30.8 11.4 12.4 21.1
Reference 0.757 0.416 0.676 0.436 0.318 0.647 1.19 0.293 0.471 0.585
5.2 24.2
Model A* 95% CI
p Value
1.512, 1.966 1.184, 1.694
<0.001 <0.001
0.945, 1.149
0.41
0.662, 0.345, 0.524, 0.265, 0.076, 0.562, 0.948, 0.126, 0.403, 0.397,
0.865 0.500 0.871 0.719 1.324 0.746 1.494 0.680 0.552 0.863
<0.001 <0.001 0.003 0.001 0.12 <0.001 0.13 0.004 <0.001 0.007
Not included in the model
aOR
Model By 95% CI
p Value
Reference 1.696 1.097
1.244, 2.312 0.733, 1.642
<0.001 0.65
0.884, 1.359
0.41
1.096 Reference Reference 0.464 0.433 0.565 0.337 2.247 0.492 0.861 0.201 0.472 0.483
0.343, 0.260, 0.340, 0.109, 0.399, 0.361, 0.523, 0.024, 0.335, 0.216,
Reference 5.558
4.071, 7.589
<0.001
0.626 0.721 0.936 1.036 12.658 0.671 1.418 1.676 0.665 1.078
<0.001 0.001 0.027 0.06 0.36 <0.001 0.56 0.14 < 0.001 0.076
22.5 17.0 14.2 16.8
Reference 0.656 0.556 0.714
0.595, 0.723 0.310, 0.997 0.581, 0.877
<0.001 0.049 0.001
Reference 0.935 0.477 0.71
0.734, 1.191 0.129, 1.763 0.450, 1.119
0.58 0.27 0.14
19.0 25.2
Reference 1.131
0.972, 1.316
0.11
Reference 0.865
0.638, 1.172
0.35
17.3 21.6
Reference 1.314
1.190, 1.452
<0.001
Reference 1.178
0.956, 1.451
0.12
*The model includes age, sex, payer of health service, procedure site, admission/trauma status, obesity status, and hospital teaching status. y The model includes the same factors as in model A, as well as race and MORPI. These factors were not included in model A because of more than 50% missing response values. No. of cases ¼ 652, and no. of controls ¼ 2,935. aOR, adjusted odds ratio; BMI, body mass index; GI, gastrointestinal; MORPI, major operating room procedure indicator; RFB, retained foreign bodies.
RFB was significantly higher compared with controls ($26,678.00 $769.69 vs $12,648.00 $192.80, p < 0.001) (Table 5). The differential cost significance remained, regardless of the complication status, procedure type, or hospital teaching status (p < 0.001 for all); however, higher monetary averages were observed for complicated cases, non-abdominopelvic procedures, and for patients managed in teaching hospitals (Table 5) (Fig. 2).
DISCUSSION In this analysis, we demonstrated that most reported RFB cases occurred in procedures that involved the abdominopelvis, the cardiovascular system, female reproductive organs, and the musculoskeletal system. However, in comparing the different surgical sites with controls, most procedures demonstrated lower risk than gastrointestinal procedures do. The distribution pattern for sites of RFB identified in this study is similar to those
Al-Qurayshi et al
Vol. 220, No. 4, April 2015
Retained Foreign Bodies: Risk and Outcomes
755
Table 3. Weighted Adjusted Odds Ratio of a Retained Foreign Body after a Procedure for Selected Factors Stratified by the Procedure Region Factor
Abdominopelvic region Age, y <36 36e65 >65 Sex Male Female MORPI Not reported Reported Admission/trauma status Elective/nontrauma Nonelective/nontrauma Elective/trauma Nonelective/trauma Obesity (BMI 30 kg/m2) No Yes Hospital teaching status Nonteaching Teaching Non-abdominopelvic region Age, y <36 36e65 >65 Sex Male Female MORPI Not reported Reported Admission/trauma status Elective/nontrauma Nonelective/nontrauma Elective/trauma Nonelective/trauma Obesity (BMI 30 kg/m2) No Yes Hospital teaching status Nonteaching Teaching
% RFB
aOR
Model A* 95% CI
10.3 28.9 29.3
Reference 3.143 3.032
2.709, 3.647 2.314, 3.974
<0.001 <0.001
28.6 17.4
1.182 Reference
1.021, 1.367
0.025
3.7 27.5
p Value
aOR
Reference 2.970 2.057
Not included in the model
Model By 95% CI
p Value
2.090, 4.221 1.164, 3.634
<0.001 0.013
0.865, 1.624
0.29
4.397, 10.301
<0.001
0.63
0.878
0.693, 1.247 Low sample size 0.265, 2.916
0.83
1.185 Reference Reference 6.730
21.8 17.5 50.6 36.9
Reference 0.749 2.635 1.697
0.663, 0.848 0.377, 18.414 0.938, 3.068
19.2 29.9
Reference 1.453
1.151, 1.834
0.002
Reference 1.144
0.750, 1.744
0.534
17.7 22.0
Reference 1.319
1.161, 1.498
<0.001
Reference 1.203
0.905, 1.599
0.2028
20.0 21.0 17.2
Reference 0.976 0.788
0.806, 1.181 0.624, 0.994
0.80 0.044
Reference 0.745 0.477
0.491, 1.132 0.285, 0.797
0.17 0.005
20.0 18.5
1.115 Reference
0.984, 1.265
0.09
0.770, 1.345
0.90
Reference 3.878
2.544, 5.910
< 0.001
7.1 21.7
<0.001 0.33 0.08
Not included in the model
Reference 0.93
1.018 Reference
23.2 16.6 11.8 15.2
Reference 0.67 0.458 0.623
0.590, 0.761 0.242, 0.868 0.500, 0.776
<0.001 0.017 <0.001
Reference 1.069 0.295 0.655
0.787, 1.453 0.067, 1.306 0.403, 1.067
0.67 0.11 0.09
18.8 22.7
Reference 1.167
0.971, 1.402
0.10
Reference 0.833
0.559, 1.243
0.37
16.9 21.2
Reference 1.298
1.140, 1.478
<0.001
Reference 1.109
0.837, 1.470
0.47
*The model includes age, sex, payer of health service, procedure site, admission/trauma status, obesity status, and hospital teaching status. y The model includes the same factors as in model A, as well as race and MORPI. These factors were not included in model A because of more than 50% missing response values. No. of cases ¼ 652, and No. of controls ¼ 2,935. aOR, adjusted odds ratio; BMI, body mass index; MORPI, major operating room procedure indicator; RFB, retained foreign bodies.
756
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
Table 4. Weighted Adjusted Odds Ratio of Hospitals with 2 or More Incidents of a Retained Foreign Body after a Procedure Characteristic
Hospital region Northeast Midwest South West Hospital size (no. of beds) Small Medium Large Hospital location Rural Urban Hospital teaching status Nonteaching Teaching
% Recurrent RFB
aOR*
95% CI
p Value
4.8 2.7 3.8 3.4
1.69 1.278, 2.235 <0.001 Reference 1.663 1.286, 2.151 <0.001 1.412 1.049, 1.902 0.023
2.8 2.6 4.2
Reference 0.808 0.567, 1.153 1.395 1.034,1.883
0.24 0.029
2.2 3.8
Reference 1.205 0.812, 1.788
0.36
2.2 5.1
Reference 2.451 1.995, 3.011 <0.001
*The model includes: hospital region, hospital bed size, hospital location, and hospital teaching status. aOR, adjusted odds ratio; RFB, retained foreign bodies.
previously described in the literature,2,5,6,38 further validating the representativeness of our sample. We showed that demographic factors are related to the risk of RFB. Men were more likely to have unintentionally retained objects after abdominopelvic procedures. Age demonstrated an unexpected association based on anatomy, such that for abdominopelvic operations, older patients (36 years old or more) had a higher risk, while for operations outside of the abdominopelvic region, younger patients (less than 36 years) had an increased risk of RFB. These findings have not been shown in previous studies, possibly due to limited sample sizes or the fact that age was treated as a continuous variable in other analyses, which could miss the association that was found here by categorizing age into groups.2,20,38,44 We observed that obesity was associated with a higher risk of RFB, particularly for abdominopelvic procedures. Two of the 3 largest case-control studies available to date also identified high BMI as a risk factor,2,20 while in a metaanalysis that involved all 3 studies, no significant association between BMI and RFB was seen.39 Our finding potentially concurs with both because we identified obesity as a risk factor in abdominopelvic interventions only, but not in the whole sample. The previously mentioned studies2,20,39 did not provide anatomic stratification to allow further verification of this hypothesis. Furthermore, despite the belief that a higher risk for RFB exists in emergency situations, this study
J Am Coll Surg
demonstrated that elective and nontraumatic admissions are more likely to result in RFB, potentially reflecting a manifestation of the stringent trauma management protocols that are followed throughout the United States. These findings were more evident for interventions outside of the abdominopelvic region, and agree with similar results from the previously mentioned meta-analysis, which showed that emergent cases were not associated with a higher risk of RFB.39 Regardless of the site of intervention; the presence of a major operating room procedure represented a recognizable risk factor. The risk of increased frequency of RFB associated with hospital characteristics has not been investigated previously. However, in this study, we showed that hospital characteristics that were associated with more than 1 event of RFB per year were teaching status and large hospitals (according to number of beds). In regard to regions throughout the United States, hospitals located in the Midwest demonstrated the lowest prevalence of recurrent RFB compared with the other regions. Results from previous case-control studies were inconsistent with regard to the risk of RFB as it relates to the presence of a trainee or resident in the operating room.2,20,38 However, no significant difference in risk was observed in the previously mentioned meta-analysis.39 As for the association of consequences with RFB, postoperative complications were twice as likely to be observed in patients with RFB compared with controls. Complications due to retained objects have been described frequently in previous investigations; however, the lack of comparison groups in most of these studies limits the significance of their findings.2,17,38,45 The costs associated with the management of patients whose cases were complicated by RFB were significantly higher compared with controls, regardless of procedure site or hospital setting. Shah and Lander46 previously reported similar findings with respect to hospital charges for pediatric operations at the national level. One of the main strengths of this study is its sample size, the largest to date, over an extended study period of 7 years, and that it was performed using a weighted analysis that reflects the most accurate and representative estimates at the national level. Additionally, the design and methodology shed light on underexplored risk factors that demonstrate significant influence on a larger scale compared with previous studies. Major limitations of this study are the administrative nature of the database and its lack of information regarding the operative environment and other details. This limited the comparability of our analysis with previously published studies. Taking into account that RFB are suggested to be under-reported events, this may place more restraint on establishing the full impact of RFB. A
Al-Qurayshi et al
Vol. 220, No. 4, April 2015
Retained Foreign Bodies: Risk and Outcomes
757
Table 5. Risk of Postoperative Complication by Type and the Cost of Health Services per Patient in Relation to Retained Foreign Body Status Type of complication
1 or more, % Cardiovascular, % Pulmonary, % Acute renal failure, % Bleeding, % Infection/sepsis, % Wound complication, % Shock, % Cystitis, % Cost, $, mean (95%CI) Overall Absence of complications Presence of complications Abdominopelvic procedures Non-abdominopelvic procedures Nonteaching hospital Teaching hospital
Retained foreign body status Casesz (n ¼ 3,045) Controlsz (n ¼ 12,592)
30.0 1.2 10.9 6.6 8.2 12.6 0.4 0.6 0.2
15.7 0.4 7.3 4.2 4.6 3.7 0.1 0.1 0.2
Risk* y 95% CI ,
aOR
2.491 2.683 1.707 1.558 1.784 3.838 4.973 11.549 1.438
2.253, 1.690, 1.479, 1.314, 1.518, 3.297, 1.897, 4.249, 0.536,
2.753 4.260 1.970 1.847 2.096 4.467 13.039 31.390 3.857
p Value
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.001 <0.001 0.47
(25,168.56e28,187.23) 12,648.00 (12,269.73e13,025.90) (17,710.74e19,987.55) 10,018.00 (9,739.91e10,295.63) (40,791.79e48,386.59) 26,553.00 (24,955.08e28,151.82) (18,449.11e22,180.46) 7,615.65 (7,263.17e7,968.13)
<0.001x <0.001x <0.001x <0.001x
32,999.00 (30,892.63e35,106.10) 17,360.00 (16,803.05e17,916.80) 20,859.00 (19,403.62e22,314.20) 11,044.00 (10,626.56e11,460.97) 31,134.00 (28,749.12e33,518.32) 14,290.00 (13,661.91e14,918.63)
<0.001x <0.001x <0.001x
26,678.00 18,849.00 44,589.00 20,315.00
*Multivariate logistic regression model for the risk of complications comparing cases to control. y The model includes: retained foreign body status, age, sex, payer of health service, procedure region, admission/trauma status, obesity status, Charlson Comorbidity Index Score, hospital region, hospital bed size, and hospital teaching status. z Column percent, for each type of complication the percent complement the absence of that particular complication. x Student’s t-test. aOR, adjusted odds ratio.
further limitation of the study is the assumption that RFB were identified during the same hospital admission as the original procedure; the NIS database does not allow tracking of patients after they are discharged. We were not able to identify patients who were diagnosed with
RFB on readmissions; however, previous studies showed that the majority of foreign bodies are identified before patient discharge.17 Current literature review revealed a wide range of policies and protocols that could be implemented to target
Figure 2. Means and 95% confidence intervals of the cost per patient based on the retained foreign body status and postoperative complications.
758
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
medical errors resulting in RFB2-7,10-19; those methods included administrative policies that regulate the operative environment and protocols that ensure an efficient interpersonal communication.2-7,10-19 Although those methods were successful in decreasing the incidence of RFB, none of them was perfectly immune to unintentionally retained objects,20 which led to the assumption by some investigators that other risk factors are still unexplored.20 Based on this concept, our study was devised in an effort to provide a risk profile of RFB from the national perspective. The observed findings warrant further exploration of the identified risk groups and the adoption of targeted approaches aimed at limiting RFB incidence in high risk population.
CONCLUSIONS In summary, RFB are not rare events, and they portend serious clinical and economic impact. This study presents a new risk profile assessment of RFB at the national level that is different from previously recognized risk factors and is tailored to a wide range of surgical specialties. Our findings enable a targeted approach to identify high-risk populations based on fundamental factors, such as patients’ clinical and demographic attributes. The discovery that elective and nontraumatic admissions are more likely to result in RFB compared with urgent trauma procedures, potentially the result of stringent trauma management protocols, suggests that protocols (checklists) for other procedures may reduce the risk of RFB. Finally, the recognized risk patterns associated with certain hospital characteristics provide an unprecedented administrative tool that may be used in screening hospitals with expected higher risk of medical errors and that mandate further protocol development by health policy makers. Author Contributions Study conception and design: Al-Qurayshi, Hauch, Slakey, Kandil Acquisition of data: Kandil Analysis and interpretation of data: Al-Qurayshi Drafting of manuscript: Al-Qurayshi, Hauch, Slakey, Kandil Critical revision: Slakey, Kandil REFERENCES 1. National Quality Forum (NQF). Serious reportable events in Healthcared2011 update: A consensus report 2011. Available at: http://www.doh.wa.gov/Portals/1/Documents/2900/NQF 2011Update.pdf. 2. Gawande AA, Studdert DM, Orav EJ, et al. Risk factors for retained instruments and sponges after surgery. N Engl J Med 2003;348:229e235.
J Am Coll Surg
3. Cima RR, Kollengode A, Garnatz J, et al. Incidence and characteristics of potential and actual retained foreign object events in surgical patients. J Am Coll Surg 2008;207:80e87. 4. Cima RR, Kollengode A, Storsveen AS, et al. A multidisciplinary team approach to retained foreign objects. Jt Comm J Qual Patient Saf 2009;35:123e132. 5. Hariharan D, Lobo DN. Retained surgical sponges, needles and instruments. Ann R Coll Surg Engl 2013;95:87e92. 6. Stawicki SP, Evans DC, Cipolla J, et al. Retained surgical foreign bodies: A comprehensive review of risks and preventive strategies. Scand J Surg 2009;98:8e17. 7. Gibbs VC. Thinking in three’s: Changing surgical patient safety practices in the complex modern operating room. World J Gastroenterol 2012;18:6712e6719. 8. Wang CF, Cook CH, Whitmill ML, et al. Risk factors for retained surgical foreign bodies: A meta-analysis. OPUS 12 Scientist 2009;2:21e27. 9. Ulucay T, Dizdar MG, SunayYavuz M, Asirdizer M. The importance of medico-legal evaluation in a case with intraabdominal gossypiboma. Forensic Sci Int 2010;198:e15e18. 10. Steelman VM, Alasagheirin MH. Assessment of radiofrequency device sensitivity for the detection of retained surgical sponges in patients with morbid obesity. Arch Surg 2012;147:955e960. 11. Chen Q, Rosen AK, Cevasco M, et al. Detecting patient safety indicators: How valid is “foreign body left during procedure” in the veterans health administration? J Am Coll Surg 2011; 212:977e983. 12. Riley R, Manias E, Polglase A. Governing the surgical count through communication interactions: Implications for patient safety. Qual Saf Health Care 2006;15:369e374. 13. Korcok M. Tracking itinerant patients and surgical sponges. CMAJ 2009;180:E14e15. 14. Colbert S, Jackson M, Turner M, Brennan PA. Reducing the risk of retained throat packs after surgery. Br J Oral Maxillofac Surg 2012;50:680e681. 15. Mahran MA, Toeima E, Morris EP. The recurring problem of retained swabs and instruments. Best Pract Res Clin Obstet Gynaecol 2013;27:489e495. 16. McIntyre LK, Jurkovich GJ, Gunn ML, Maier RV. Gossypiboma: Tales of lost sponges and lessons learned. Arch Surg 2010;145:770e775. 17. Stawicki SP, Cook CH, Anderson HL 3rd, et al. Natural history of retained surgical items supports the need for team training, early recognition, and prompt retrieval. Am J Surg 2014;208:65e72. 18. Fabian CE. Electronic tagging of surgical sponges to prevent their accidental retention. Surgery 2005;137:298e301. 19. Egorova NN, Moskowitz A, Gelijns A, et al. Managing the prevention of retained surgical instruments: What is the value of counting? Ann Surg 2008;247:13e18. 20. Stawicki SP, Moffatt-Bruce SD, Ahmed HM, et al. Retained surgical items: A problem yet to be solved. J Am Coll Surg 2013;216:15e22. 21. Prasad S, Krishnan A, Limdi J, Patankar T. Imaging features of gossypiboma: Report of two cases. J Postgrad Med 1999;45: 18e19. 22. Gostling J, McKain ES, Dumont S, Stephenson BM. Lost ‘swab’ at inguinal hernioplasty under local anaesthesia. Ann R Coll Surg Engl 2006;88:161e162. 23. Almeida AL, Cavalcante VM, Teixeira MD, Freitas GR. Pacemaker electrode misplaced in the left ventricle. Arq Bras Cardiol 2010;95:e83e87.
Vol. 220, No. 4, April 2015
24. Nayyar R, Chavda S, Singh P, Gupta NP. Missed diagnosis of anterior urethral valve complicated with a foreign body: A cause for concern. Afr J Paediatr Surg 2011;8:89e91. 25. Koul PA, Mufti SA, Khan UH, Jan RA. Intrathoracic gossypiboma causing intractable cough. Interact Cardiovasc Thorac Surg 2012;14:228e230. 26. Nemati MH. Mediastinal gossypiboma simulating a malignant tumour. Interact Cardiovasc Thorac Surg 2012;15:783e785. 27. Celkan MA, Bayatli K. A bulldog clamp that was forgotten during a coronary artery bypass operation 8 years ago. Interact Cardiovasc Thorac Surg 2012;15:777e778. 28. Hajj-Chahine J, Jayle C, Tomasi J, Corbi P. eComment. Late presentation of thoracic textiloma. Interact Cardiovasc Thorac Surg 2012;15:785. 29. Shen HP, Tseng CJ, Lin LY, et al. Retroperitoneal gossypiboma 25 years after abdominal hysterectomy. Taiwan J Obstet Gynecol 2012;51:460e462. 30. Carroll MI, Ahanchi SS, Kim JH, Panneton JM. Endovascular foreign body retrieval. J Vasc Surg 2013;57:459e463. 31. Sankar V, Shakeel M, Keh S, Ah-See KW. A case of a ‘lost’ nasogastric tube. J Laryngol Otol 2012;126:1296e1298. 32. Jaiswal A, Habib M, Tanwar YS. Neglected foreign body in contralateral limb in a traumatic transfemoral amputee- radiographs can be misleading. Chin J Traumatol 2013;16:61e64. 33. Gil-Romea I, Valcarreres-Rivera MP, Palacios-Gasos P, et al. Postoperative intra-abdominal foreign body resemblance to neoplasm. Report of a case and review of the bibliography. Cir Cir 2013;81:148e152. 34. Adenekan AT, Onakpoya UU, Faponle AF, Olateju SO. Unrecognised guide wire migration during internal jugular cannulation and its retrieval- a case report. Niger Postgrad Med J 2013;20:63e65. 35. Xu J, Wang H, Song ZW, et al. Foreign body retained in liver long after gauze packing. World J Gastroenterol 2013;19: 3364e3368. 36. Olusanya AA, Akinmoladun VI. Orbito-antro-cervical foreign body impaction: Reminder of a CT scan and ultrasonography pitfall. Afr J Med Med Sci 2013;42:189e192. 37. Massimiliano PA, Massimo PS. Retained intra-abdominal surgical instrument: A rare condition of acute abdomen. ANZ J Surg 2010;80:758. 38. Lincourt AE, Harrell A, Cristiano J, et al. Retained foreign bodies after surgery. J Surg Res 2007;138:170e174. 39. Moffatt-Bruce SD, Cook CH, Steinberg SM, Stawicki SP. Risk factors for retained surgical items: A meta-analysis and proposed risk stratification system. J Surg Res 2014;190: 429e436. 40. Healthcare Cost and Utilization Project. Overview of national inpatient sample. Available at: http://www.hcup-us.ahrq.gov/ nisoverview.jsp. Updated 2013. Accessed September 1, 2013. 41. Healthcare Cost and Utilization Project. Clinical classifications software for ICD-9-CM. Available at: http://hcup-us.ahrq.gov/ toolssoftware/ccs/ccs.jsp. Updated 2014. Accessed December 22, 2014. 42. Healthcare Cost and Utilization Project. NIS description of data elements. Available at: http://hcup-us.ahrq.gov/db/ nation/nis/nisdde.jsp. Updated 2008. Accessed December 22, 2014. 43. 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;43:1130e1139.
Al-Qurayshi et al
Discussion
759
44. Moffatt-Bruce SD, Ellison EC, Anderson HL 3rd, et al. Intravascular retained surgical items: A multicenter study of risk factors. J Surg Res 2012;178:519e523. 45. Gonzalez-Ojeda A, Rodriguez-Alcantar DA, ArenasMarquez H, et al. Retained foreign bodies following intraabdominal surgery. Hepatogastroenterology 1999;46:808e812. 46. Shah RK, Lander L. Retained foreign bodies during surgery in pediatric patients: A national perspective. J Pediatr Surg 2009; 44:738e742.
Discussion DR DOUGLAS TYLER (Galveston, TX): As the authors highlight in their presentation, retained foreign bodies are “never” events that continue to happen with distressing frequency. While the authors’ work highlights some new association and risk factors associated with retained foreign bodies, at the end of the day, we are still trying to find some guidance as to how these events can be prevented. Many institutions have implemented mandatory x-rays or wands that can detect chipped embedded sponges, but system errors leading to retained foreign bodies still occur despite these interventions. This relates to my questions for the authors. First, are there any data about the systems that were in place in the facilities that were studied in this presentation that could give us some insight into the incidence of retained foreign bodies that occurred with these systems in place? For example, is there any way of knowing whether implementation of a system had a positive effect? Are there any data that will allow us, over time, to see what happened so we can see trends in retained foreign bodies? Second, do you have any access to the Veterans Affairs Surgical Quality Improvement Program (VASQIP) database, which actually does capture this information prospectively? It may provide not only additional information regarding potential risk factors but also facilitate the studies on the efficacy of target interventions. Finally, what would you recommend now for operating rooms as a way of minimizing or eliminating retained foreign bodies, as you started to allude to in your conclusion? Is there really a best practice out there with some of the interventions we currently have available? DR MICHAEL S NUSSBAUM (Jacksonville, FL): This paper addresses an issue that continues to challenge surgeons despite the development of many supposed preventive tools and technologies. From approximately 8 million hospital stays and about 1,000 hospitals, the authors identified 3,045 cases of retained foreign bodies and present a new risk profile that is different from previously recognized risk factors. The majority of incidents were reported after gastrointestinal interventions. Risk of retained foreign body was higher in teaching hospitals. Surprisingly, patients admitted with traumatic injuries had no higher risk when undergoing abdominopelvic procedures and were actually at lower risk for procedures unrelated to abdominopelvic operations. Obesity and older age were associated with a
MD, MBA,
Douglas P Slakey,
MD, MPH, FACS,
Retained foreign bodies (RFB) after operative interventions are linked to an increased risk of morbidity and mortality, and represent a medico-legal liability. We aimed to identify associated risk factors and outcomes related to iatrogenic RFB in the United States. STUDY DESIGN: A cross-sectional analysis was performed on all interventions that resulted in a secondary diagnosis of RFB in the Nationwide Inpatient Sample (NIS) from 2003 to 2009. Comparative controls were randomly selected from patients who underwent similar procedures. RESULTS: We identified 3,045 cases of RFB, and 12,592 controls were included. The majority of incidents, 968 (31.8%), were reported after gastrointestinal interventions. Risk of RFB was higher in teaching hospitals (odds ratio [OR] 1.31, 95% CI [1.19, 1.45], p < 0.001). For abdominopelvic procedures, patients admitted with traumatic injuries did not demonstrate a higher risk of RFB compared with electively admitted patients (OR 1.70, 95% CI [0.94, 3.07], p ¼ 0.08). However, for procedures unrelated to abdominopelvic surgery, patients admitted for trauma had a lower risk (OR 0.62, 95% CI [0.50, 0.78], p < 0.001). Obesity (BMI 30 kg/m2) and older age (65 years) were significantly associated with a higher risk only for abdominopelvic procedures (p < 0.01 for both). Retained foreign bodies were associated with a higher average cost of health services ($26,678.00 $769.69 vs $12,648.00 $192.80, p < 0.001). CONCLUSIONS: Retained foreign bodies have unfavorable and nationally tangible clinical and economic outcomes. The risk profile for RFB at the national level seems to demonstrate an association with demographic and clinical factors including nature of the procedure, type of admission, and trauma status. Teaching hospitals are associated with a higher risk. Targeted efforts toward identified high-risk populations are needed to avoid these morbid and costly complications. (J Am Coll Surg 2015;220:749e759. 2015 by the American College of Surgeons)
BACKGROUND:
inpatient operations in the United States.2-4 Sponges are the most commonly retained items compared with needles and instruments.3,5-7 Regardless of the nature of the foreign body, the consequences of left-behind objects include grievous health implications to patients, indefensible litigations, and irreparable damage to the reputations of health care professionals and institutions.5,6,8-10 Despite the theoretically preventable nature of RFB, tackling this medical error represents a true challenge to health care policy makers.6 Several standardized protocols and regulations have been devised to mitigate this issue, like manual and automated sponge and instrument counting, radiographic screening, and goals for improvement of communication among the surgical team; however, no method has proven to be completely successful in preventing RFB.2-7,10-19 Moreover, it is hypothesized that the
Retained foreign bodies (RFB) after interventional procedures are recognized by the National Quality Forum as preventable medical errors, listed as “never events” in their updated 2011 report Serious Reportable Events In Healthcare.1 The true incidence of RFB is still elusive; it is estimated that it ranges from 1 in every 1,000 to 1,500 abdominal operations and 1 in every 8,000 to 18,000 Disclosure Information: Nothing to disclose. Presented at the Southern Surgical Association 126th Annual Meeting, Palm Beach, FL, November 30eDecember 3, 2014. Received December 7, 2014; Accepted December 9, 2014. From the Department of Surgery, Tulane University School of Medicine, New Orleans, LA. Correspondence address: Emad Kandil, MD, FACS, Department of Surgery, Tulane University School of Medicine, 1430 Tulane Ave, Room 8510 (Box SL-22), New Orleans, LA 70112. email: [email protected]
ª 2015 by the American College of Surgeons Published by Elsevier Inc.
749
http://dx.doi.org/10.1016/j.jamcollsurg.2014.12.015 ISSN 1072-7515/14
750
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
Abbreviations and Acronyms
BMI ¼ body mass index HCUP ¼ Healthcare Cost and Utilization Project ICD-9 ¼ International Classification of Disease, 9th Revision MORPI ¼ major operating room procedure indicator NIS ¼ Nationwide Inpatient Sample OR ¼ odds ratio RFB ¼ retained foreign bodies
reason behind the continuing occurrence of this complication is that some risk factors remain unexplored.20 Studies that have addressed RFB consist predominately of case reports and case series.6,21-37 The main hindrance in conducting larger scale studies is the lack of considerable aggregate data, due in part to the natural history of RFB in which an unknown number of events can go unnoticed.8,17 Additionally, some investigators have proposed that hospitals are inclined toward under-reporting such incidents because of their critical nature.5,6,8,9,38 This becomes obvious when reviewing the literature because most of the larger analytic studies depend on data from malpractice lawsuits.2,6 To date, 3 relatively large casecontrol studies, with RFB sample sizes of less than 60, have examined the risk factors for RFB after a variety of surgical interventions.2,20,38 Among these studies, discrepancies exist in identifying persistent risk factors.2,20,38 However, in a recent meta-analysis that included the previously mentioned studies, several factors exhibited significant risk for unintentional RFB, including intraoperative blood loss of more than 500 mL, longer duration of operation, involvement of more than 1 subprocedure or surgical team, unexpected intraoperative factors, and finally, lack of or incorrect surgical count.39 In this study, we aimed to assess the frequency of RFB at the national level and across all disciplines, as opposed to previous studies that have focused on certain institutions’ experiences and specific fields. We also aimed to identify the risk factors for and the clinical and economic outcomes associated with RFB in the United States by using data from the Nationwide Inpatient Sample (NIS) for the period of 2003 to 2009.
METHODS This study is a cross-sectional analysis using the Nationwide Inpatient Sample (NIS) database for the years 2003 to 2009. The NIS is part of the Healthcare Cost and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality (AHRQ). This is the largest all-payer inpatient care database that
J Am Coll Surg
is publicly available in the United States. It contains data from approximately 8 million hospital stays from about 1,000 hospitals sampled to approximate a 20% stratified sample of US community hospitals.40 Although NIS data are available for the period between 1988 to 2010 up to the time of the data analysis, the data elements and the sampling methodologies were evolving and changing over the years, which put a technical restraint on longitudinal analyses, and for which the HCUP is prompting investigators to be cautious with data before and after the period included in this analysis.40 For these coding precautions, we opted to include the data analysis between 2003 and 2009 to perform a thorough, accurate, and unjeopardized investigation. The Clinical Classifications Software (CCS) was used to identify the primary procedures of interest41; the International Classification of Disease, 9th Revision (ICD-9) was used in defining the other parameters of the study. This software is a diagnosis and procedure categorization scheme developed by HCUP and based on the ICD-9; it provides a smaller number of clinically meaningful categories that are more useful for presenting descriptive statistics than are individual ICD-9 codes.41 The database was surveyed for patients who developed a secondary diagnosis of RFB (ICD-9: 998.5) during their hospital stay. Then, based on their primary procedure, controls were randomly selected from those who underwent the same procedures performed within the same year. The main risk factors for RFB that were assessed included: Age less than 36, 36 to 65, and greater than 65 years old; sex; site of procedure, determined based on the involved system by the primary procedure: gastrointestinal tract/spleen, cardiovascular system, respiratory system, nervous system, head and neck, eye, musculoskeletal system, urinary system, breast, female reproductive system, and male reproductive system; NIS major operating room procedure indicator (MORPI): not reported vs reported42; admission/trauma status: elective/nontrauma, nonelective/nontrauma, elective/trauma, and nonelective/trauma; body mass index (BMI) 30 kg/ m2; and hospital teaching status: nonteaching vs teaching. However, in assessing the above risk factors, more than 50% of the records were missing response values for race and MORPI. In order not to lose the power of the study, 2 multivariate logistic regression models were performed, primary model “A” excluding race and MORPI, and a secondary model “B” including race and MORPI. The results of both models are provided. Hospital characteristics that were examined for the risk of more than 1 incidence of RFB per year were: Hospital region: Northeast, Midwest, West, and South; hospital
Al-Qurayshi et al
Vol. 220, No. 4, April 2015
Table 1.
Retained Foreign Bodies: Risk and Outcomes
751
Descriptive Statistics of the Study Population in Relation to Outcomes Related to Retained Foreign Body
Variable
Age, y <36 36e65 >65 Sex Male Female Race White Black Hispanic Asian/Pacific Islander Native American Other Annual household income, $ <39,000 39,000e47,999 48,000e62,999 >62,999 Service payer Medicare Medicaid Private/HMO Self-pay No charges Other Procedure site/system GI tract/spleen Cardiovascular Respiratory Nervous Head and neck Eye Musculoskeletal Urinary Breast Female reproductive Male reproductive MORPI Not reported Reported Procedure region Non-abdominopelvic Abdominopelvic Admission/trauma status Elective/nontrauma Nonelective/nontrauma
% Study population (n ¼ 15,637)
Retained foreign body status % Cases (n ¼ 3,045) % Controls (n ¼ 12,592)
p Value*
30.6 38.2 31.2
19.3 47.7 33.0
33.3 35.9 30.8
<0.001
36.4 63.6
41.9 58.1
35.1 64.9
<0.001
69.2 11.7 12.6 2.5 0.6 3.4
72.4 11.8 10.3 2.0 0.7 2.8
68.4 11.7 13.1 2.7 0.6 3.5
<0.001
26.6 26.2 24.3 22.9
26.7 25.9 25.3 22.0
26.5 26.3 24.1 23.1
0.45
34.2 16.9 40.8 4.1 0.4 3.7
37.3 12.5 41.9 3.8 0.5 4.0
33.4 18.0 40.5 4.2 0.4 3.6
<0.001
24.2 15.5 8.0 3.0 0.8 0.1 16.1 2.9 0.5 27.6 1.2
31.8 18.3 4.9 3.4 0.6 0.1 17.1 4.6 0.3 17.6 1.3
22.3 14.8 8.8 3.0 0.8 0.1 15.9 2.5 0.5 30.1 1.2
<0.001
30.8 69.2
8.7 91.3
35.8 64.2
<0.001
51.2 48.8
50.6 49.4
51.4 48.6
0.44
44.7 49.5
51.7 43.4
43.0 51.0 (Continued)
752
Table 1.
Al-Qurayshi et al
J Am Coll Surg
Retained Foreign Bodies: Risk and Outcomes
Continued
Variable
Elective/trauma Nonelective/trauma CCIS Low: 0e1 Medium-low: 2e3 Intermediate: 4e5 High: 6 Obesity (BMI 30 kg/m2) No Yes In-hospital mortality No Yes Hospital region Northeast Midwest South West Hospital size (no. of beds) Small Medium Large Hospital location Rural Urban Hospital teaching status Nonteaching Teaching
% Study population (n ¼ 15,637)
Retained foreign body status % Cases (n ¼ 3,045) % Controls (n ¼ 12,592)
p Value*
0.6 5.2
0.5 4.5
0.7 5.3
<0.001
96.3 3.7 0.1 0.0
95.8 4.2 0.0 0.0
96.4 3.6 0.1 0.0
0.18
93.0 7.0
90.9 9.1
93.5 6.5
<0.001
97.1 2.9
96.6 3.4
97.3 2.7
0.05
19.5 23.7 36.2 20.6
19.9 23.1 35.0 22.1
19.5 23.8 36.5 20.3
0.29
10.9 24.2 64.9
11.0 23.4 65.6
10.9 24.4 64.7
0.60
10.0 90.0
10.3 89.7
9.9 90.1
0.55
49.2 50.8
43.6 56.4
50.5 49.5
<0.001
*Chi-square test. BMI, body mass index; CCIS, Charlson Comorbidity Index Score; GI, gastrointestinal; MORPI, major operating room procedure indicator.
size (number of beds): small, medium, and large42; hospital location: urban vs rural; and hospital teaching status: teaching hospital vs nonteaching. The outcomes of RFB were assessed by examining 2 factors: Absence or presence of 1 or more postoperative complications, classified into cardiovascular complications, pulmonary complications, acute renal failure, bleeding, infection/sepsis, wound complications, shock, and cystitis; and average cost of the health services on the hospital per patient. Other secondary independent factors that were assessed as potential confounders were patient race (white, black, Hispanic, Asian/Pacific Islander, Native American, and other), a modification of the Charlson Comorbidity Index Score (CCIS) (none, mild, moderate, severe),43 inpatient death, annual household income (quartile classification: <$39,000; $39,000 to $47,999;
$48,000 to $62,999; and >$62,999), main payer of health service (Medicare, Medicaid, private insurance, self-pay, no charges, and other). Statistical analysis used weighted data reflecting a national estimate. The records’ weights are available in the NIS data and calculated based on the stratification variables that were used in sampling methodology. These variables include hospital geographic region, urban or rural location, teaching status, hospital size (number of beds), and ownership. Cross-tabulation and chi-square tests were used to examine the association between each of the independent factors and the outcome of interest. Factors with significant association were considered possible confounders and were included in multivariate logistic regression models. Multivariate logistic regression models were used to calculate the odds ratio (OR) and 95% CI.
Vol. 220, No. 4, April 2015
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
753
Figure 1. Frequency of retained foreign bodies according to the type of procedure, 2003 to 2009.
Student’s t-test was used to test for the differences in cost. Stratification by the site of the procedure (abdominopelvic procedures vs non-abdominopelvic procedures) and hospital teaching status was conducted in the assessment of the factors of interest. Significance level was set as (a ¼ 0.05). All data analyses were performed using SAS 9.2 for Windows (SAS Institute Inc).
RESULTS There were 3,045 cases of RFB identified, and 12,592 controls were included (Table 1). A control:case ratio of at least 4:1 for each procedure code was included. The mean age of the study population was 50.9 years (0.3 years [SEM]), with the majority being white (69.2%), female (63.6%), and with private health insurance (40.8%) (Table 1). There were 102 RFB patients who died during their hospital stay, compared with 364 controls (p > 0.05). The most common procedure categories of RFB were for gastrointestinal procedures (968 cases; 31.8%), while interventions for the cardiovascular system, female reproductive organs, and musculoskeletal system followed with nearly identical frequencies of 554, 538, and 523 cases, respectively (Fig. 1). Overall, younger patients (less than 36 years) had a lower risk of RFB compared with older age groups (p < 0.001) (Table 2). On stratification by procedure type (Table 3), the same relationship was observed for abdominopelvic operations. In non-abdominopelvic operations, the younger age group had a marginally higher risk of RFB compared with patients older than 65 years (20% vs 17.2%, p ¼ 0.044). Males were more likely to have retained objects only for procedures in the abdominopelvic region (OR 1.18, 95% CI [1.02, 1.37], p ¼ 0.025) (Table 3). Similarly, obesity was a significant risk factor for abdominopelvic operations (OR 1.45, 95% CI [1.15, 1.83], p ¼ 0.002) (Tables 2 and 3).
Abdominopelvic procedures were significantly associated with a higher risk of RFB compared with other sites (p < 0.05 for all) (Table 2). However, there was no significant difference in the prevalence of RFB between gastrointestinal and urinary system procedures (OR 1.19, 95%CI [0.95, 1.49], p ¼ 0.13). On the other hand, patients who underwent major operating room procedures (MORPI) were at least 4 times more likely to have a retained object, regardless of the surgery site (OR 5.56, 95% CI [4.07, 7.59], p < 0.001). In general, and for interventions in the nonabdominopelvic region specifically, elective admissions and nontraumatic presentations were associated with a higher prevalence of RFB compared with nonelective and traumatic presentations (p < 0.05) (Table 2). However, for patients who required abdominopelvic procedures, there was no significant difference among patients sustaining trauma, regardless of their admission type, compared with those with elective or nontrauma status (p > 0.05). With respect to hospital characteristics, teaching hospitals had a higher incidence of RFB compared with nonteaching facilities (OR 1.31, 95% CI [1.19, 1.45], p < 0.001) (Tables 1 and 2). Furthermore, the risk of more than 1 incident of RFB per year was significantly higher in teaching and large hospitals (p < 0.05) (Table 4). The lowest incidence of recurrent RFB was in hospitals located in the Midwest, while the remaining regions (Northeast, South, and West), all had significantly higher incidences compared with the Midwest (p < 0.05) (Table 4). Cases of unintentionally retained objects, compared with controls, were twice as likely to have 1 or more postoperative complications (OR 2.49, 95% CI [2.25, 2.75], p < 0.001) (Table 5). The average cost (SEM) of hospitalization for a patient having an
754
Table 2.
Al-Qurayshi et al
J Am Coll Surg
Retained Foreign Bodies: Risk and Outcomes
Weighted Adjusted Odds Ratio of a Retained Foreign Body after a Procedure for Selected Risk Factors
Factor
Age, y <36 36e65 >65 Sex Male Female Procedure site/system GI tract/spleen Cardiovascular Respiratory Nervous Head and neck Eye Musculoskeletal Urinary Breast Female reproductive Male reproductive MORPI Not reported Reported Admission/trauma status Elective/nontrauma Nonelective/nontrauma Elective/trauma Nonelective/trauma Obesity (BMI 30 kg/m2) No Yes Hospital teaching status Nonteaching Teaching
% RFB
aOR
12.3 24.3 20.6
Reference 1.724 1.416
22.4 17.8
1.042 Reference
25.6 22.9 11.9 21.8 15.4 10.4 20.6 30.8 11.4 12.4 21.1
Reference 0.757 0.416 0.676 0.436 0.318 0.647 1.19 0.293 0.471 0.585
5.2 24.2
Model A* 95% CI
p Value
1.512, 1.966 1.184, 1.694
<0.001 <0.001
0.945, 1.149
0.41
0.662, 0.345, 0.524, 0.265, 0.076, 0.562, 0.948, 0.126, 0.403, 0.397,
0.865 0.500 0.871 0.719 1.324 0.746 1.494 0.680 0.552 0.863
<0.001 <0.001 0.003 0.001 0.12 <0.001 0.13 0.004 <0.001 0.007
Not included in the model
aOR
Model By 95% CI
p Value
Reference 1.696 1.097
1.244, 2.312 0.733, 1.642
<0.001 0.65
0.884, 1.359
0.41
1.096 Reference Reference 0.464 0.433 0.565 0.337 2.247 0.492 0.861 0.201 0.472 0.483
0.343, 0.260, 0.340, 0.109, 0.399, 0.361, 0.523, 0.024, 0.335, 0.216,
Reference 5.558
4.071, 7.589
<0.001
0.626 0.721 0.936 1.036 12.658 0.671 1.418 1.676 0.665 1.078
<0.001 0.001 0.027 0.06 0.36 <0.001 0.56 0.14 < 0.001 0.076
22.5 17.0 14.2 16.8
Reference 0.656 0.556 0.714
0.595, 0.723 0.310, 0.997 0.581, 0.877
<0.001 0.049 0.001
Reference 0.935 0.477 0.71
0.734, 1.191 0.129, 1.763 0.450, 1.119
0.58 0.27 0.14
19.0 25.2
Reference 1.131
0.972, 1.316
0.11
Reference 0.865
0.638, 1.172
0.35
17.3 21.6
Reference 1.314
1.190, 1.452
<0.001
Reference 1.178
0.956, 1.451
0.12
*The model includes age, sex, payer of health service, procedure site, admission/trauma status, obesity status, and hospital teaching status. y The model includes the same factors as in model A, as well as race and MORPI. These factors were not included in model A because of more than 50% missing response values. No. of cases ¼ 652, and no. of controls ¼ 2,935. aOR, adjusted odds ratio; BMI, body mass index; GI, gastrointestinal; MORPI, major operating room procedure indicator; RFB, retained foreign bodies.
RFB was significantly higher compared with controls ($26,678.00 $769.69 vs $12,648.00 $192.80, p < 0.001) (Table 5). The differential cost significance remained, regardless of the complication status, procedure type, or hospital teaching status (p < 0.001 for all); however, higher monetary averages were observed for complicated cases, non-abdominopelvic procedures, and for patients managed in teaching hospitals (Table 5) (Fig. 2).
DISCUSSION In this analysis, we demonstrated that most reported RFB cases occurred in procedures that involved the abdominopelvis, the cardiovascular system, female reproductive organs, and the musculoskeletal system. However, in comparing the different surgical sites with controls, most procedures demonstrated lower risk than gastrointestinal procedures do. The distribution pattern for sites of RFB identified in this study is similar to those
Al-Qurayshi et al
Vol. 220, No. 4, April 2015
Retained Foreign Bodies: Risk and Outcomes
755
Table 3. Weighted Adjusted Odds Ratio of a Retained Foreign Body after a Procedure for Selected Factors Stratified by the Procedure Region Factor
Abdominopelvic region Age, y <36 36e65 >65 Sex Male Female MORPI Not reported Reported Admission/trauma status Elective/nontrauma Nonelective/nontrauma Elective/trauma Nonelective/trauma Obesity (BMI 30 kg/m2) No Yes Hospital teaching status Nonteaching Teaching Non-abdominopelvic region Age, y <36 36e65 >65 Sex Male Female MORPI Not reported Reported Admission/trauma status Elective/nontrauma Nonelective/nontrauma Elective/trauma Nonelective/trauma Obesity (BMI 30 kg/m2) No Yes Hospital teaching status Nonteaching Teaching
% RFB
aOR
Model A* 95% CI
10.3 28.9 29.3
Reference 3.143 3.032
2.709, 3.647 2.314, 3.974
<0.001 <0.001
28.6 17.4
1.182 Reference
1.021, 1.367
0.025
3.7 27.5
p Value
aOR
Reference 2.970 2.057
Not included in the model
Model By 95% CI
p Value
2.090, 4.221 1.164, 3.634
<0.001 0.013
0.865, 1.624
0.29
4.397, 10.301
<0.001
0.63
0.878
0.693, 1.247 Low sample size 0.265, 2.916
0.83
1.185 Reference Reference 6.730
21.8 17.5 50.6 36.9
Reference 0.749 2.635 1.697
0.663, 0.848 0.377, 18.414 0.938, 3.068
19.2 29.9
Reference 1.453
1.151, 1.834
0.002
Reference 1.144
0.750, 1.744
0.534
17.7 22.0
Reference 1.319
1.161, 1.498
<0.001
Reference 1.203
0.905, 1.599
0.2028
20.0 21.0 17.2
Reference 0.976 0.788
0.806, 1.181 0.624, 0.994
0.80 0.044
Reference 0.745 0.477
0.491, 1.132 0.285, 0.797
0.17 0.005
20.0 18.5
1.115 Reference
0.984, 1.265
0.09
0.770, 1.345
0.90
Reference 3.878
2.544, 5.910
< 0.001
7.1 21.7
<0.001 0.33 0.08
Not included in the model
Reference 0.93
1.018 Reference
23.2 16.6 11.8 15.2
Reference 0.67 0.458 0.623
0.590, 0.761 0.242, 0.868 0.500, 0.776
<0.001 0.017 <0.001
Reference 1.069 0.295 0.655
0.787, 1.453 0.067, 1.306 0.403, 1.067
0.67 0.11 0.09
18.8 22.7
Reference 1.167
0.971, 1.402
0.10
Reference 0.833
0.559, 1.243
0.37
16.9 21.2
Reference 1.298
1.140, 1.478
<0.001
Reference 1.109
0.837, 1.470
0.47
*The model includes age, sex, payer of health service, procedure site, admission/trauma status, obesity status, and hospital teaching status. y The model includes the same factors as in model A, as well as race and MORPI. These factors were not included in model A because of more than 50% missing response values. No. of cases ¼ 652, and No. of controls ¼ 2,935. aOR, adjusted odds ratio; BMI, body mass index; MORPI, major operating room procedure indicator; RFB, retained foreign bodies.
756
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
Table 4. Weighted Adjusted Odds Ratio of Hospitals with 2 or More Incidents of a Retained Foreign Body after a Procedure Characteristic
Hospital region Northeast Midwest South West Hospital size (no. of beds) Small Medium Large Hospital location Rural Urban Hospital teaching status Nonteaching Teaching
% Recurrent RFB
aOR*
95% CI
p Value
4.8 2.7 3.8 3.4
1.69 1.278, 2.235 <0.001 Reference 1.663 1.286, 2.151 <0.001 1.412 1.049, 1.902 0.023
2.8 2.6 4.2
Reference 0.808 0.567, 1.153 1.395 1.034,1.883
0.24 0.029
2.2 3.8
Reference 1.205 0.812, 1.788
0.36
2.2 5.1
Reference 2.451 1.995, 3.011 <0.001
*The model includes: hospital region, hospital bed size, hospital location, and hospital teaching status. aOR, adjusted odds ratio; RFB, retained foreign bodies.
previously described in the literature,2,5,6,38 further validating the representativeness of our sample. We showed that demographic factors are related to the risk of RFB. Men were more likely to have unintentionally retained objects after abdominopelvic procedures. Age demonstrated an unexpected association based on anatomy, such that for abdominopelvic operations, older patients (36 years old or more) had a higher risk, while for operations outside of the abdominopelvic region, younger patients (less than 36 years) had an increased risk of RFB. These findings have not been shown in previous studies, possibly due to limited sample sizes or the fact that age was treated as a continuous variable in other analyses, which could miss the association that was found here by categorizing age into groups.2,20,38,44 We observed that obesity was associated with a higher risk of RFB, particularly for abdominopelvic procedures. Two of the 3 largest case-control studies available to date also identified high BMI as a risk factor,2,20 while in a metaanalysis that involved all 3 studies, no significant association between BMI and RFB was seen.39 Our finding potentially concurs with both because we identified obesity as a risk factor in abdominopelvic interventions only, but not in the whole sample. The previously mentioned studies2,20,39 did not provide anatomic stratification to allow further verification of this hypothesis. Furthermore, despite the belief that a higher risk for RFB exists in emergency situations, this study
J Am Coll Surg
demonstrated that elective and nontraumatic admissions are more likely to result in RFB, potentially reflecting a manifestation of the stringent trauma management protocols that are followed throughout the United States. These findings were more evident for interventions outside of the abdominopelvic region, and agree with similar results from the previously mentioned meta-analysis, which showed that emergent cases were not associated with a higher risk of RFB.39 Regardless of the site of intervention; the presence of a major operating room procedure represented a recognizable risk factor. The risk of increased frequency of RFB associated with hospital characteristics has not been investigated previously. However, in this study, we showed that hospital characteristics that were associated with more than 1 event of RFB per year were teaching status and large hospitals (according to number of beds). In regard to regions throughout the United States, hospitals located in the Midwest demonstrated the lowest prevalence of recurrent RFB compared with the other regions. Results from previous case-control studies were inconsistent with regard to the risk of RFB as it relates to the presence of a trainee or resident in the operating room.2,20,38 However, no significant difference in risk was observed in the previously mentioned meta-analysis.39 As for the association of consequences with RFB, postoperative complications were twice as likely to be observed in patients with RFB compared with controls. Complications due to retained objects have been described frequently in previous investigations; however, the lack of comparison groups in most of these studies limits the significance of their findings.2,17,38,45 The costs associated with the management of patients whose cases were complicated by RFB were significantly higher compared with controls, regardless of procedure site or hospital setting. Shah and Lander46 previously reported similar findings with respect to hospital charges for pediatric operations at the national level. One of the main strengths of this study is its sample size, the largest to date, over an extended study period of 7 years, and that it was performed using a weighted analysis that reflects the most accurate and representative estimates at the national level. Additionally, the design and methodology shed light on underexplored risk factors that demonstrate significant influence on a larger scale compared with previous studies. Major limitations of this study are the administrative nature of the database and its lack of information regarding the operative environment and other details. This limited the comparability of our analysis with previously published studies. Taking into account that RFB are suggested to be under-reported events, this may place more restraint on establishing the full impact of RFB. A
Al-Qurayshi et al
Vol. 220, No. 4, April 2015
Retained Foreign Bodies: Risk and Outcomes
757
Table 5. Risk of Postoperative Complication by Type and the Cost of Health Services per Patient in Relation to Retained Foreign Body Status Type of complication
1 or more, % Cardiovascular, % Pulmonary, % Acute renal failure, % Bleeding, % Infection/sepsis, % Wound complication, % Shock, % Cystitis, % Cost, $, mean (95%CI) Overall Absence of complications Presence of complications Abdominopelvic procedures Non-abdominopelvic procedures Nonteaching hospital Teaching hospital
Retained foreign body status Casesz (n ¼ 3,045) Controlsz (n ¼ 12,592)
30.0 1.2 10.9 6.6 8.2 12.6 0.4 0.6 0.2
15.7 0.4 7.3 4.2 4.6 3.7 0.1 0.1 0.2
Risk* y 95% CI ,
aOR
2.491 2.683 1.707 1.558 1.784 3.838 4.973 11.549 1.438
2.253, 1.690, 1.479, 1.314, 1.518, 3.297, 1.897, 4.249, 0.536,
2.753 4.260 1.970 1.847 2.096 4.467 13.039 31.390 3.857
p Value
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.001 <0.001 0.47
(25,168.56e28,187.23) 12,648.00 (12,269.73e13,025.90) (17,710.74e19,987.55) 10,018.00 (9,739.91e10,295.63) (40,791.79e48,386.59) 26,553.00 (24,955.08e28,151.82) (18,449.11e22,180.46) 7,615.65 (7,263.17e7,968.13)
<0.001x <0.001x <0.001x <0.001x
32,999.00 (30,892.63e35,106.10) 17,360.00 (16,803.05e17,916.80) 20,859.00 (19,403.62e22,314.20) 11,044.00 (10,626.56e11,460.97) 31,134.00 (28,749.12e33,518.32) 14,290.00 (13,661.91e14,918.63)
<0.001x <0.001x <0.001x
26,678.00 18,849.00 44,589.00 20,315.00
*Multivariate logistic regression model for the risk of complications comparing cases to control. y The model includes: retained foreign body status, age, sex, payer of health service, procedure region, admission/trauma status, obesity status, Charlson Comorbidity Index Score, hospital region, hospital bed size, and hospital teaching status. z Column percent, for each type of complication the percent complement the absence of that particular complication. x Student’s t-test. aOR, adjusted odds ratio.
further limitation of the study is the assumption that RFB were identified during the same hospital admission as the original procedure; the NIS database does not allow tracking of patients after they are discharged. We were not able to identify patients who were diagnosed with
RFB on readmissions; however, previous studies showed that the majority of foreign bodies are identified before patient discharge.17 Current literature review revealed a wide range of policies and protocols that could be implemented to target
Figure 2. Means and 95% confidence intervals of the cost per patient based on the retained foreign body status and postoperative complications.
758
Al-Qurayshi et al
Retained Foreign Bodies: Risk and Outcomes
medical errors resulting in RFB2-7,10-19; those methods included administrative policies that regulate the operative environment and protocols that ensure an efficient interpersonal communication.2-7,10-19 Although those methods were successful in decreasing the incidence of RFB, none of them was perfectly immune to unintentionally retained objects,20 which led to the assumption by some investigators that other risk factors are still unexplored.20 Based on this concept, our study was devised in an effort to provide a risk profile of RFB from the national perspective. The observed findings warrant further exploration of the identified risk groups and the adoption of targeted approaches aimed at limiting RFB incidence in high risk population.
CONCLUSIONS In summary, RFB are not rare events, and they portend serious clinical and economic impact. This study presents a new risk profile assessment of RFB at the national level that is different from previously recognized risk factors and is tailored to a wide range of surgical specialties. Our findings enable a targeted approach to identify high-risk populations based on fundamental factors, such as patients’ clinical and demographic attributes. The discovery that elective and nontraumatic admissions are more likely to result in RFB compared with urgent trauma procedures, potentially the result of stringent trauma management protocols, suggests that protocols (checklists) for other procedures may reduce the risk of RFB. Finally, the recognized risk patterns associated with certain hospital characteristics provide an unprecedented administrative tool that may be used in screening hospitals with expected higher risk of medical errors and that mandate further protocol development by health policy makers. Author Contributions Study conception and design: Al-Qurayshi, Hauch, Slakey, Kandil Acquisition of data: Kandil Analysis and interpretation of data: Al-Qurayshi Drafting of manuscript: Al-Qurayshi, Hauch, Slakey, Kandil Critical revision: Slakey, Kandil REFERENCES 1. National Quality Forum (NQF). Serious reportable events in Healthcared2011 update: A consensus report 2011. Available at: http://www.doh.wa.gov/Portals/1/Documents/2900/NQF 2011Update.pdf. 2. Gawande AA, Studdert DM, Orav EJ, et al. Risk factors for retained instruments and sponges after surgery. N Engl J Med 2003;348:229e235.
J Am Coll Surg
3. Cima RR, Kollengode A, Garnatz J, et al. Incidence and characteristics of potential and actual retained foreign object events in surgical patients. J Am Coll Surg 2008;207:80e87. 4. Cima RR, Kollengode A, Storsveen AS, et al. A multidisciplinary team approach to retained foreign objects. Jt Comm J Qual Patient Saf 2009;35:123e132. 5. Hariharan D, Lobo DN. Retained surgical sponges, needles and instruments. Ann R Coll Surg Engl 2013;95:87e92. 6. Stawicki SP, Evans DC, Cipolla J, et al. Retained surgical foreign bodies: A comprehensive review of risks and preventive strategies. Scand J Surg 2009;98:8e17. 7. Gibbs VC. Thinking in three’s: Changing surgical patient safety practices in the complex modern operating room. World J Gastroenterol 2012;18:6712e6719. 8. Wang CF, Cook CH, Whitmill ML, et al. Risk factors for retained surgical foreign bodies: A meta-analysis. OPUS 12 Scientist 2009;2:21e27. 9. Ulucay T, Dizdar MG, SunayYavuz M, Asirdizer M. The importance of medico-legal evaluation in a case with intraabdominal gossypiboma. Forensic Sci Int 2010;198:e15e18. 10. Steelman VM, Alasagheirin MH. Assessment of radiofrequency device sensitivity for the detection of retained surgical sponges in patients with morbid obesity. Arch Surg 2012;147:955e960. 11. Chen Q, Rosen AK, Cevasco M, et al. Detecting patient safety indicators: How valid is “foreign body left during procedure” in the veterans health administration? J Am Coll Surg 2011; 212:977e983. 12. Riley R, Manias E, Polglase A. Governing the surgical count through communication interactions: Implications for patient safety. Qual Saf Health Care 2006;15:369e374. 13. Korcok M. Tracking itinerant patients and surgical sponges. CMAJ 2009;180:E14e15. 14. Colbert S, Jackson M, Turner M, Brennan PA. Reducing the risk of retained throat packs after surgery. Br J Oral Maxillofac Surg 2012;50:680e681. 15. Mahran MA, Toeima E, Morris EP. The recurring problem of retained swabs and instruments. Best Pract Res Clin Obstet Gynaecol 2013;27:489e495. 16. McIntyre LK, Jurkovich GJ, Gunn ML, Maier RV. Gossypiboma: Tales of lost sponges and lessons learned. Arch Surg 2010;145:770e775. 17. Stawicki SP, Cook CH, Anderson HL 3rd, et al. Natural history of retained surgical items supports the need for team training, early recognition, and prompt retrieval. Am J Surg 2014;208:65e72. 18. Fabian CE. Electronic tagging of surgical sponges to prevent their accidental retention. Surgery 2005;137:298e301. 19. Egorova NN, Moskowitz A, Gelijns A, et al. Managing the prevention of retained surgical instruments: What is the value of counting? Ann Surg 2008;247:13e18. 20. Stawicki SP, Moffatt-Bruce SD, Ahmed HM, et al. Retained surgical items: A problem yet to be solved. J Am Coll Surg 2013;216:15e22. 21. Prasad S, Krishnan A, Limdi J, Patankar T. Imaging features of gossypiboma: Report of two cases. J Postgrad Med 1999;45: 18e19. 22. Gostling J, McKain ES, Dumont S, Stephenson BM. Lost ‘swab’ at inguinal hernioplasty under local anaesthesia. Ann R Coll Surg Engl 2006;88:161e162. 23. Almeida AL, Cavalcante VM, Teixeira MD, Freitas GR. Pacemaker electrode misplaced in the left ventricle. Arq Bras Cardiol 2010;95:e83e87.
Vol. 220, No. 4, April 2015
24. Nayyar R, Chavda S, Singh P, Gupta NP. Missed diagnosis of anterior urethral valve complicated with a foreign body: A cause for concern. Afr J Paediatr Surg 2011;8:89e91. 25. Koul PA, Mufti SA, Khan UH, Jan RA. Intrathoracic gossypiboma causing intractable cough. Interact Cardiovasc Thorac Surg 2012;14:228e230. 26. Nemati MH. Mediastinal gossypiboma simulating a malignant tumour. Interact Cardiovasc Thorac Surg 2012;15:783e785. 27. Celkan MA, Bayatli K. A bulldog clamp that was forgotten during a coronary artery bypass operation 8 years ago. Interact Cardiovasc Thorac Surg 2012;15:777e778. 28. Hajj-Chahine J, Jayle C, Tomasi J, Corbi P. eComment. Late presentation of thoracic textiloma. Interact Cardiovasc Thorac Surg 2012;15:785. 29. Shen HP, Tseng CJ, Lin LY, et al. Retroperitoneal gossypiboma 25 years after abdominal hysterectomy. Taiwan J Obstet Gynecol 2012;51:460e462. 30. Carroll MI, Ahanchi SS, Kim JH, Panneton JM. Endovascular foreign body retrieval. J Vasc Surg 2013;57:459e463. 31. Sankar V, Shakeel M, Keh S, Ah-See KW. A case of a ‘lost’ nasogastric tube. J Laryngol Otol 2012;126:1296e1298. 32. Jaiswal A, Habib M, Tanwar YS. Neglected foreign body in contralateral limb in a traumatic transfemoral amputee- radiographs can be misleading. Chin J Traumatol 2013;16:61e64. 33. Gil-Romea I, Valcarreres-Rivera MP, Palacios-Gasos P, et al. Postoperative intra-abdominal foreign body resemblance to neoplasm. Report of a case and review of the bibliography. Cir Cir 2013;81:148e152. 34. Adenekan AT, Onakpoya UU, Faponle AF, Olateju SO. Unrecognised guide wire migration during internal jugular cannulation and its retrieval- a case report. Niger Postgrad Med J 2013;20:63e65. 35. Xu J, Wang H, Song ZW, et al. Foreign body retained in liver long after gauze packing. World J Gastroenterol 2013;19: 3364e3368. 36. Olusanya AA, Akinmoladun VI. Orbito-antro-cervical foreign body impaction: Reminder of a CT scan and ultrasonography pitfall. Afr J Med Med Sci 2013;42:189e192. 37. Massimiliano PA, Massimo PS. Retained intra-abdominal surgical instrument: A rare condition of acute abdomen. ANZ J Surg 2010;80:758. 38. Lincourt AE, Harrell A, Cristiano J, et al. Retained foreign bodies after surgery. J Surg Res 2007;138:170e174. 39. Moffatt-Bruce SD, Cook CH, Steinberg SM, Stawicki SP. Risk factors for retained surgical items: A meta-analysis and proposed risk stratification system. J Surg Res 2014;190: 429e436. 40. Healthcare Cost and Utilization Project. Overview of national inpatient sample. Available at: http://www.hcup-us.ahrq.gov/ nisoverview.jsp. Updated 2013. Accessed September 1, 2013. 41. Healthcare Cost and Utilization Project. Clinical classifications software for ICD-9-CM. Available at: http://hcup-us.ahrq.gov/ toolssoftware/ccs/ccs.jsp. Updated 2014. Accessed December 22, 2014. 42. Healthcare Cost and Utilization Project. NIS description of data elements. Available at: http://hcup-us.ahrq.gov/db/ nation/nis/nisdde.jsp. Updated 2008. Accessed December 22, 2014. 43. 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;43:1130e1139.
Al-Qurayshi et al
Discussion
759
44. Moffatt-Bruce SD, Ellison EC, Anderson HL 3rd, et al. Intravascular retained surgical items: A multicenter study of risk factors. J Surg Res 2012;178:519e523. 45. Gonzalez-Ojeda A, Rodriguez-Alcantar DA, ArenasMarquez H, et al. Retained foreign bodies following intraabdominal surgery. Hepatogastroenterology 1999;46:808e812. 46. Shah RK, Lander L. Retained foreign bodies during surgery in pediatric patients: A national perspective. J Pediatr Surg 2009; 44:738e742.
Discussion DR DOUGLAS TYLER (Galveston, TX): As the authors highlight in their presentation, retained foreign bodies are “never” events that continue to happen with distressing frequency. While the authors’ work highlights some new association and risk factors associated with retained foreign bodies, at the end of the day, we are still trying to find some guidance as to how these events can be prevented. Many institutions have implemented mandatory x-rays or wands that can detect chipped embedded sponges, but system errors leading to retained foreign bodies still occur despite these interventions. This relates to my questions for the authors. First, are there any data about the systems that were in place in the facilities that were studied in this presentation that could give us some insight into the incidence of retained foreign bodies that occurred with these systems in place? For example, is there any way of knowing whether implementation of a system had a positive effect? Are there any data that will allow us, over time, to see what happened so we can see trends in retained foreign bodies? Second, do you have any access to the Veterans Affairs Surgical Quality Improvement Program (VASQIP) database, which actually does capture this information prospectively? It may provide not only additional information regarding potential risk factors but also facilitate the studies on the efficacy of target interventions. Finally, what would you recommend now for operating rooms as a way of minimizing or eliminating retained foreign bodies, as you started to allude to in your conclusion? Is there really a best practice out there with some of the interventions we currently have available? DR MICHAEL S NUSSBAUM (Jacksonville, FL): This paper addresses an issue that continues to challenge surgeons despite the development of many supposed preventive tools and technologies. From approximately 8 million hospital stays and about 1,000 hospitals, the authors identified 3,045 cases of retained foreign bodies and present a new risk profile that is different from previously recognized risk factors. The majority of incidents were reported after gastrointestinal interventions. Risk of retained foreign body was higher in teaching hospitals. Surprisingly, patients admitted with traumatic injuries had no higher risk when undergoing abdominopelvic procedures and were actually at lower risk for procedures unrelated to abdominopelvic operations. Obesity and older age were associated with a