Coexisting Systemic Infections in Patients Who Present With a Fall

Author’s Accepted Manuscript Coexisting Systemic Infections in Patients Who Present with a Fall Alex Blair, Farrin A. Manian

www.elsevier.com

PII: DOI: Reference:

S0002-9629(16)30615-2 http://dx.doi.org/10.1016/j.amjms.2016.11.010 AMJMS327

To appear in: The American Journal of the Medical Sciences Received date: 8 June 2016 Revised date: 1 November 2016 Accepted date: 4 November 2016 Cite this article as: Alex Blair and Farrin A. Manian, Coexisting Systemic Infections in Patients Who Present with a Fall, The American Journal of the Medical Sciences, http://dx.doi.org/10.1016/j.amjms.2016.11.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Coexisting Systemic Infections in Patients Who Present with a Fall

Alex Blair MD, MS*, Farrin A. Manian MD, MPH*† Harvard Medical School*, Department of Medicine, Massachusetts General Hospital†, Boston, MA Key words: fall, infections, sepsis, bacteremia Running title: Coexisting infections in falls Corresponding author: Farrin A. Manian, MD, MPH Department of Medicine, Massachusetts General Hospital, 503B Staniford, Boston, MA 02114 [email protected] Phone: 617.643.9258 Fax:617.724.9428

Presented in part at IDWeek (poster #813), October 9, 2015, San Diego, CA.

1

Abstract Background: Although the causes of falls are legion, infectious disease-related factors are not commonly reported in the published literature. We investigated the characteristics of patients presenting to the hospital because of a fall and who were subsequently found to have a coexisting systemic infection (CSI). Methods: This was a retrospective study performed at Massachusetts General Hospital, using the electronic data base of adult patients receiving care during the period January 1, 2000, through December 31, 2014. Cases were initially screened by using billing codes for “fall”, ”sepsis”, “bacteremia”, and “systemic inflammatory response syndrome” (SIRS). Evaluable patients had documented CSI in the setting of a fall. Results: Of 161 evaluable patients, 84 (52.2%) were female. The mean age was 75. 2 years (range 35102 years, median 78 years). Fall was considered “mechanical” (e.g. tripped by a rug) in 106 (65.8%) cases, with 126 (78.3%) patients living at home. SIRS criteria were met on initial healthcare encounters of 66 (40.1%) patients. Urinary and lower respiratory tract infections were the most common infectious disease conditions (71[44.1%], and 37[23.0%] cases, respectively). Bacteremia was present in 64 (39.8%) cases. Staphylococcus aureus was the most common cause of bacteremia (21 cases, 31.3% of bloodstream isolates). CSI was not initially suspected by providing clinicians in 64 (39.8%) patients. Conclusion: Falls associated with CSIs are often considered “mechanical” in nature, and frequently fail to meet the SIRS criteria on initial presentation. Aside from its commonly-recognized causes, falls may be an atypical manifestation of a systemic infection.

2

INTRODUCTION Falls are a leading cause of injury and death, afflicting about one-third of adults over 65 years of age annually1. Although causes of falls are legion, infectious diseases-related factors have received relatively little attention1-4. The few studies that have implicated infections as a risk factor for falls have involved primarily elderly patients with dementia and urinary tract infections (UTIs) in institutionalized settings3,5,6, limiting their generalizability to a relatively small segment of the population. We hypothesized that the spectrum of patients who fall in the setting of coexisting systemic infections (CSIs) may be much broader than previously reported and wished to better characterize the profile of such patients as well as their CSIs. METHODS This was a retrospective study performed at Massachusetts General Hospital (MGH), a teaching tertiary care institution in Boston. The computerized Research Patient Data Registry (RPRDR), a central electronic data repository containing health records of adult patients (≥ 18 years of age) enrolled at MGH was employed by using the International Classification of Diseases, 9th revision, (ICD-9) coding, using the following search terms : “Fall” (E880, E8844, E8843, E8846, and E8859), and “bacteremia” (7907), “systemic inflammatory response syndrome”(SIRS) or sepsis (9959) or “unspecified septicemia” (0389). The study period spanned from January 1, 2000 through December 31, 2014. Patients were considered eligible only if there was documentation of a chief complaint of a recent fall less than 7 days prior to their presentation, were admitted to the hospital, and were found to have a CSI within 48 hours of their presentation. Hospital-related falls were excluded. In the case of more than 1 evaluable fall with CSI in a patient during the study period, only the most recent episode was considered for analysis. Bacteremia included those with primary (e.g. intravascular device-related) or secondary causes.

3

SIRS was defined as the presence of 2 or more of the following commonly accepted criteria: temperature ≥38° C or <36° C, heart rate >90/min, respiration rate > 20/min, PaCO2<32 mm/hg, and WBC< 4000/µL or > 12,000/µL, or immature band forms >10% (7). CSI was defined based on clinical, laboratory, and radiographic assessment by the providing physicians (s) and requirement for systemic antimicrobial therapy. Acute bacterial skin and soft tissue infections (ABSSTIs) consisted of cellulitis, infected wounds or furunculosis. “Mechanical” fall was defined as an accidental fall caused by tripping over an object (e.g. a rug), slipping, or loss of balance in the absence of reported syncope (i.e. transient loss of consciousness) or seizure. “Syncopal” falls were associated with transient loss of consciousness. Patient demographics, clinical data, details of the fall, and the providing physicians’ overall assessment and management at the time of the initial encounter based on emergency department or admission notes were recorded on a standard study form for individual cases. The study was approved by the Institutional Review Board of MGH. Fisher’s exact and Student T tests were used to compare categorical and continuous data, respectively. Statistical analysis was performed utilizing Statistica (StatSoft Inc, Tulsa, OK) and Instat (GraphPad Software, Inc., San Diego, CA) softwares, with P<0.05 considered statistically significant. Results Of 406 patients whose diagnoses initially matched the specified ICD-9 codes, 245 were excluded due to: lack of documentation of recent fall as the primary reason for seeking medical care (176 cases), fall outside of the study period (30 cases), CSI not documented (29 cases), and fall occurring in a hospital setting (10 cases). Of the remaining 161 cases, 84 (52.2%) were female. The mean age was 75. 2 years (range 35-102 years, median 78 years). The age distribution of patients by deciles is shown (Figure). Patients less than 70 years of age accounted for 41 (25.5%) of cases. 4

At the time of the fall, 126 (78.3%) patients lived at home, and 15 (9.3%) patients resided at an extended care facility; the place of residence of the remaining 20 patients was unclear. Falls were considered mechanical in 106 (65.8%), syncopal in 12 (7.5%), and unclassified in 43 (26.7%) patients. Characteristics of patients by selected comorbidities, reported symptoms preceding the fall, physical examination findings, fall-related fractures, and in-hospital mortality by age group (less than 70 years vs ≥ 70 years) are shown (Table 1). There was a significant gender difference in age distribution with the majority of men falling in the younger age group, while the majority of women were represented by the older age group. One or more pre-existing medical conditions were present in 126 (78.3%) patients; cardiovascular disease, diabetes mellitus and malignancy were the most common. Chronic alcoholism was significantly more prevalent in the less than 70 year age group. Alzheimer’s dementia was reported in 8.1% of patients. One or more symptoms (ie, weakness or lethargy, dizziness or lightheadedness, mental status changes, or fever) during the days leading to the fall were reported in 53 (32.9%) patients with the majority of cases represented by the older age group (83.0 %). SIRS criteria (≥2) was met in 66 (40.1%) patient. Alcohol intoxication was reported in 6.8% of patients, with the majority of cases represented by the younger age group (72.7%). There was no significant difference in the rates of fever or abnormal temperature, mental status changes, or hypotension between the 2 age groups. Fallrelated fractures were documented in 18.6% of patients with hip or femur fractures present in 8.7% of cases. No significant difference in the distribution of fractures was found between the 2 age groups. In-hospital mortality was 18% without significant difference in its distribution between the 2 age groups. There was no also significant association between gender, and pre-existing comorbidities, fractures or in-hospital mortality (data not shown).

5

Table 2 shows the breakdown of CSIs by specific infectious disease-related diagnoses. UTI was the most frequent (44.1%), followed by lower respiratory tract infection (LRTI) (23%), sepsis of unclear source (21.1%) and endocarditis (5.6%). ABSSTIs accounted for 3.7% of all CSIs. Concurrent bacteremia was present on 1 or more occasions in all categories of CSIs and was reported in 30.9% of UTIs. Overall, 64 (39.8%) patients had documented bacteremia with no vascular device-related infections reported. . UTI was significantly more likely to account for CSIs among females: 45 (53.6%) vs 26 (33.8%) for males (O.R. 2.2, 95% C.I. 1.2-4.3, P=0.02). Otherwise, there was no significant association between categories of infection and gender, age group, fractures, or in-hospital mortality (data not shown). Blood culture isolates in patients with bacteremia are listed (table 3). Staphylococcus aureus (31.3%), E. coli (19.4%), Klebsiella sp (7.5%). and Enterococcus sp (7.5%) were the most common pathogens. CSI was not suspected by the providing clinician(s) on initial healthcare encounter in 64 (39.8%) cases, and was not considered a possible explanation for the fall in 123 (76.4%) patients. Seventy-seven (47.8%) patients were discharged to an extended care facility, 41(25.5%) were released home, while 14 (8.7%) patients had miscellaneous discharge status (e.g. discharge against medical advice, transfer to another hospital or unclear destination). Discussion An estimated 8-45% of falls have been attributed to infections based on a limited number of studies involving primarily institutionalized elderly patients with UTI3,5,6. The results of our study, however, suggest that the spectrum of patients presenting with a fall often includes those younger than 70 years of age, patients living at home without known dementia, and frequently involve infections other than UTIs. We found that the majority of males were in the less than 70 year age group, while the majority of females were in the ≥ 70 year age group. Whether this finding is related to women generally living longer and possibly more likely to live alone and seek medical attention, or other reasons is unclear.

6

Selected symptoms such as weakness or lethargy, dizziness or lightheadedness, mental status changes, and fever antedated some falls, but were present in nearly a third (32.9%) of patients with the great majority (83%) of symptomatic patients falling in the older than 70 year age group. Whether these symptoms can be used as potential markers for falls with CSIs is unclear in the absence of a comparative control group. Of interest, despite the potentially serious nature of many CSIs (e.g. bacteremia) in our patient population, only 20% had fever or abnormal temperature on presentation and the majority (59.9 %) did not meet the SIRS definition for sepsis. Suboptimal performance of fever or SIRS criteria in detecting potentially serious infections have been previously reported in other patient populations8-10, particularly the elderly with bacteremia8. Of interest, the majority (66%) of our patients appeared to have a “mechanical fall”, a popular (perhaps oversimplified) label that may unintentionally divert attention away from other less apparent contributors to falls, including those of infectious origin.

Thus, in contrast to a previous model of falls

where acute medical illness serves as a “precipitating cause” rather than a pre-existing condition11, it is

,

possible that an acute medical illness such as CSI may serve as a pre-existing condition with

environmental hazards, such as tripping over a rug, serving as a precipitating cause instead. Further studies are needed to better delineate the relative contribution of infectious and mechanical factors in fall causation. Although UTI was the most common CSI in our patient population particularly in women, bacteremia and LRTIs were also significantly represented. In contrast, ABSSTIs, one of the most common infectious conditions among patients requiring hospitalization12, accounted for only 3.7% of CSIs. Whether this finding is related to earlier detection of signs of ABSSTI by patients or their family member before the infection becomes severe enough to lead to a fall or other factors is not clear.

7

Although urinary pathogens, such as E. coli, Klebsiella sp., and Enterococcus sp. accounted for over onethird of bacteremias, S. aureus was the most common while several other organisms not commonly originating from the urinary tract, including Streptococcus pneumoniae and anaerobes, were also represented. These findings support the varied nature of infections that may potentially play a causative role in falls. We also found that CSI was not suspected by providing clinicians in nearly 40% of cases, with even fewer clinicians considering it in their differential diagnosis of conditions that may play a causative role. Although these findings may be in part related to the paucity of signs of inflammation (e.g. lack of fever) among many of our patients, other factors such as lack of appreciation of sepsis as a potential cause of diffuse muscle weakness13,14--and therefore falls--- and a preoccupation with the mechanical aspects of the fall itself, rather than the underlying medical conditions that may lead to it, might have also played a role. Further diversion of attention away from an underlying infection may also occur in the setting of serious fall-related complications (eg, fractures or subdural hematomas) that may need urgent surgical intervention. Under such circumstances, lack of consideration of CSI in the differential diagnosis of conditions presenting with a fall may be particularly problematic not only because of the need for timely antimicrobial therapy in sepsis15-17, but also because such patients may unknowingly be subjected to the additional risk of perioperative myocardial depression, vasodilatation and hemodynamic instability caused by the anesthesia itself15,18,19. The observed 18.6% rate of fractures in our patient population was higher than the 5% rate reported in the literature for all falls 20 . Similarly the 18% rate of in-hospital mortality found in our study was higher than the 3% rate reported among a nationwide sample of patients 65 years or older hospitalized for falls in the United States21. It is unclear whether these findings reflect the adverse impact of CSI on the severity of falls and recovery from them. Without a comparative group of patients who present with

8

falls without CSI, however, the attributable risk of fractures or mortality related to CSIs cannot be determined. Several potential mechanisms may be invoked in explaining how CSIs could contribute to falls or impact their severity, including possible hypovolemia due to poor oral intake in the setting of an infection, as well as sepsis-related hemodynamic instability, confusion, and myopathy14,15,22,23. In addition to systemic factors, night time urinary urgency may also contribute to falls in patients with UTI6. Our study has several limitations worthy of emphasis. First, we did not study patients who presented with a fall in the absence of a CSI, and were thus unable to determine independent risk factors associated with CSI-related falls. Second, it is likely that our patient population represented a more severe spectrum of CSIs since we specifically targeted patients with a coexisting diagnosis of sepsis or bacteremia. Third, we could not entirely exclude the possibility of some infections arising as a result of complications (e.g. aspiration and skin trauma) from the fall itself. However, since many of our patients presented soon after their fall, and had no historical or clinical evidence to suggest such scenario (e.g. no loss of consciousness or evidence of ABSSTI on exam), this possibility seems less likely. Lastly, this was a single center study with results that may not necessarily be generalizable to centers with a different patient-mix than ours. In conclusion, the results of ours study suggest that falls may be associated with potentially serious systemic infections, including bacteremia. Inclusion of systemic infections in the differential diagnosis of conditions that may contribute or coexist with a fall should lead to their more timely diagnosis and management. Further studies are needed to identify clinical parameters that predict CSI in patients who present with a fall.

9

References 1. Ambrose AC, Paul G, Hasudorff JM. Risk factors for falls among older adults: a review of the literature. Maturitas 75;2013:51-61. 2. Tinetti ME, Kumar C. The patient who falls: “It’s always a trade-off”. JAMA 2010;303:258–66. 3. Kallin K, Jensen J, Olsson LL et al. Why the elderly fall in residential care facilities, and suggested remedies. J Fam Pract 2004;53:41–52. 4. Deandrea S, Lucenteforte E, Bravi F, et al. Risk factors for falls in community-dwelling older people: a systematic review and meta-analysis. Epidemiology 2010; 21(5):658–668. 5. Tangman S, Eriksson S. Yngve G, Lundin-Olsson L. Precipitating factors for falls among patients with dementia on a psychogeriatric ward. Int Psychogeriatr 2010; 22:641-649. 6. Rhoads J, Clayman A. Nelson S. The relationship of urinary tract infections and falls in a nursing home. Director 2007; 15:22-6. 7. Bone RC, Balk RA, Cerra FB, et al. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 1992;101:1644–55. 8. Manian FA. Fever, abnormal white blood cell count, neutrophilia, and elevated serum creactive protein in adult hospitalized patients with bacteremia. S Med J 2012;105:474-8. 9. Kaukonen KM, Bailey M, Pilcher D. et al. Systemic inflammatory response syndrome criteria in defining severe sepsis. N Engl J Med 2015; 372:1629-1638. 10. Lindvig KP, Henriksen DP, Nielsen SL, et al. How do bacteraemic patients present to the emergency department and what is the diagnostic validity of the clinical parameters;

10

temperature, C-reactive protein and systemic inflammatory response syndrome? Scand J Trauma Resusc Emerg Med 2014; 22:39. 11. Rubenstein, LZ, Josephson KR. Falls and their prevention in elderly people. What does the evidence show? Med Clin North Am 2006;90:807-23. 12. Jenkins TC, Sabel AL, Sarcone EE, et al. Skin and soft-tissue infections requiring hospitalization at an academic medical center; opportunities for antimicrobial stewardship. Clin Infect Dis 2010;51:895-903. 13. Alamdari N, Toraldo G, Aversa Z, et al. Loss of muscle strength during sepsis is in part regulated by glucocorticoids and is associated with reduced muscle fiber stiffness. Am J Physiol Regul Integr Comp Physiol 2012;303: R1090-R1099. 14. Callahan LA, Supinski GS. Sepsis-induced myopathy. Crit Care Med 2009;37: S354–S367. 15. Eissa D, Carton EG, Buggy DJ. Anaesthetic management of patients with severe sepsis. Br J Anaesth 2010;105:734-43. 16. Gauer RL. Early recognition and management of sepsis in adults: the first six hours. Am Fam Physician 2013;88(1):44-53. 17. Gaieski DF, Pines JM, Band RA, et al. Impact of time to antibiotics on survival in patients with severe sepsis or septic shock in whom early goal –directed therapy was initiated in the emergency department. Crit Care Med 2010;38:1-9. 18. Yoon SH. Concerns of the anesthesiologist: anesthetic induction in severe sepsis or septic shock patients. Korean J Anesthesiol 2012; 63:3-10. 19. Cheung CC, Martyn A, Campbell N, et al. Predictors of intraoperative hyotension and bradycardia. Am J Med 2015;128:532-538. 20. Masud T, Morris RO. Epidemiology of falls. Age and Ageing 2001;30-S4:3-7.

11

21. Moudouni DKM, Phillips CD. In-hospital mortality and unintentional falls among older adults in the United States. J Appl Gerontol 2013;32:923-35. 22. Young GB. Encephalopathy of infection and systemic inflammation. J Clin Neurophysiol 2013;30:454-461. 23. Brealey D, Brand M, Hargreaves I, et al. Association between mitochondrial dysfunction and severity and outcome of septic shock. Lancet 2002;360:219-223.

Disclosures/Disclaimers The authors have nothing to disclose. The content is solely the responsibility of the authors and does not necessarily represent the official views of Harvard Catalyst, Harvard University, its affiliate academic healthcare centers, or its corporate contributors.

Figure. Age distribution of patients with fall and coexisting systemic infection

12

Table 1. Characteristics of patients presenting with a fall and coexisting systemic infection by age group.

Variable value*

All (%)

<70 years (%)

≥70 years (%)

P

Number, 41

Number, 120

(O.R.,

77 (47.8)

27 (35.1)

50 (64.9)

0.01

84 (52.2)

14 (16.7)

70 (83.3)

Cardiovascular disease NS

72 (44.7)

15 (20.8)

57 (79.2)

Diabetes mellitus NS

37 (23.0)

12 (32.4)

25 (67.6)

Malignancy NS

35 (21.7)

7 (20)

28 (80)

Chronic renal insufficiency NS

20 (12.4)

4 (20)

16 (80)

15 (9.3)

11 (73.3)

4 (26.7)

Alzheimer’s dementia NS

13 (8.1)

1 (7.7)

12 (92.3)

COPD† NS

16 (9.9)

4 (25)

12 (75)

End-stage liver disease NS

7 (4.4)

4 (57.1)

3 (49.2)

Total number, 161 95% C.I.) Gender Male (2.7, 1.3-5.2) Female --Pre-existing diagnosis

Chronic alcoholism <0.0001 (10.6, 3.2-35.8)

13

≥1 of the above NS

126 (78.3)

30 (23.8)

96 (76.2)

Weakness or lethargy NS

26 (16.2)

4 (15.4)

22 (84.6)

Dizziness or lightheadedness NS

14 (8.7)

3 (21.4)

11 (78.6)

Mental status changes NS

9 (5.6)

1 (11.1)

8 (88.9)

Fever NS

8 (5.0)

1 (12.5)

7 (87.5)

≥1 of the above NS

53 (32.9)

9 (17.0)

44 (83.0)

0

40 (24.9)

6 (15.0)

34 (85)

1

50 (31.1)

13 (26.0)

37 (74.0)

2

43 (26.7)

13 (30.2)

30 (69.8)

3

17 (10.6)

2 (11.8)

15 (88.2)

6 (3.7)

5 (83.3)

1 (16.7)

66 (40.1)

20 (30.3)

46 (69.7)

Fever or abnormal temp.§ NS

32 (19.9)

10 (31.3)

22 (68.7)

Mental status changes NS

41 (25.5)

13 (31.7)

28 (68.3)

Reported symptoms prior to fall

Physical findings SIRS criteriaǂ

NS

NS

NS

NS 4 (16.5, 1.9-146.2) ≥2 NS

14

0.004

Hypotension¶ NS

18 (11.2)

8 (44.4)

10 (55.6)

11 (6.8)

8 (72.7)

3 (27.3)

Any fracture NS

30 (18.6)

9 (30.0)

21 (70.0)

Hip or femur fracture NS

14 (8.7)

2 (14.3)

12 (85.7)

29 (18.0)

4 (13.8)

25 (86.2)

Alcohol intoxication (9.5, 2.4-37.7)

0.0009

Fracture complicating fall

In-hospital death NS

_____________________________________________________________________________________ ________________ *Fisher’s exact and Student T tests used for categorical and continuous data, respectively; NS, not statistically significant †Chronic obstructive pulmonary disease ǂSystemic inflammatory response syndrome criteria: temperature ≥38° C or < 36° C, heart rate >90/minute, respiration rate > 20/minute, PaCO2 < 32 mm Hg, WBC < 4000/µL or > 12,000/µL, or bands > 10%; 1 or more criteria could not be determined in 6 cases. §Abnormal temperature defined as ≥38° C or < 36° C ¶Systolic blood pressure < 90 mm/Hg

15

Table 2. Coexisting systemic infections in patients presenting with a fall.

Category

Total Number (%)*

Concurrent bacteremia Number (%)†

____________________________________________________________________________________ Urinary tract infection

71 (44.1)

22 (30.9)

Lower respiratory tract infection

37 (23.0)

9 (24.3)

Sepsis of unclear source

34 (21.1)

23 (67.7)

Endocarditis

9 (5.6)

9 (100)

Acute bacterial skin and soft tissue infection

6 (3.7)

3 (50)

Cholangitis

2 (1.2)

1 (50)

Prosthetic joint infection

1 (0.6)

1 (100)

*Percentage of 161 total cases; some cases had more than 1 coexisting systemic infections † Percentage of each infection category

16

Table 3. Blood culture isolates (total number, 67 cases)*.

Pathogen

Number (%)

S. aureus†

21 (31.3)

E. coli

13 (19.4)

Klebsiella sp.

5 (7.5)

Enterococcus sp.

5 (7.5)

Β-hemolytic streptococci‡

4 (6.0)

S. pneumoniae

3 (4.5)

Pseudomonas sp.

2 (3.0)

Proteus sp.

2 (3.0)

Coagulase-negative staphylococci

2 (3.0)

Streptococcus bovis

2 (3.0)

Streptococcus mitis

1 (1.5)

Streptococcus anginosis

1 (1.5)

Enterobacter aerogenes

1 (1.5)

Clostridium sordelli

1 (1.5)

Clostridium perfringes

1 (1.5)

Veillonella sp.

1 (1.5)

Peptostreptococcus sp.

1 (1.5)

Morganella sp.

1 (1.5)

*3 blood cultures grew 2 pathogens concurrently: P. aerogenes and E. Coli, methicillin-susceptible S. aureus and Klebsiella sp., and Veillonella and Peptostreptococcus sp. †16 methicillin-susceptible, 5 methicillin-resistant isolates ‡ Includes group B (2 episodes), group A (1 episode), and group G (1 episode) streptococci.

17

Frequency (number)

Age (years)

18