N503, Week 4 Lab Sept. 28, 2004

1
N503, Week 4 LabSept. 28, 2004

• Reviewing a published quantitative research study

2
Citation

• Pitten, F. A., Kiefer, T., Buth, C., Doelken, G.,
  Kramer, A. (2003). Do cancer patients with
  chemotherapy-induced leukopenia benefit from an
  antiseptic chlorhexidine-based oral rinse? A
  double-blind, block randomized, controlled study.
  Journal of Hospital Infection 53, 283-291.

3
1. What is the research question?

• A. Restate according to Level
• 1.) Level 3 Among cancer patients who become
  leukopenic due to chemotherapy, what is the
  benefit from antiseptic rinsing with
  chlorhexidine-based mouth wash?
• B. Is it clear and logical?
• 1.) No
• a.) Restatement
• (1.) Primary question Does rinsing with a
  chlorhexidine-based mouth wash significantly
  reduce infection in cancer patients who become
  leukopenic due to chemotherapy?
• (2.) Secondary question Does rinsing with a
  chlorhexidine-based mouth wash significantly
  reduce oral bacterial count and mucositis in
  cancer patients who become leukopenic due to
  chemotherapy?

4

• C. Is the research question appropriately
  supported by the literature?
• Yes
• 1.) Because patients with myeloablative
  treatment should not brush their teeth due to the
  risk of bleeding, they have poor oral hygiene.
• 2.) Poor oral hygiene increases severity of
  mucositis and hence risk for infection.
• 3.) Chlorhexidine is a broad spectrum antiseptic
  that has been shown to reduce plaque and
  gingivitis, and risk of pneumonia (infection) in
  ICU patients.
• 4.) Conflicting findings regarding whether use
  of a chlorhexidine-based rinse prevents infection
  in cancer patients.

5

• D. What are the variables?
• 1. Independent
• a. Rinsing solution
• 2. Dependent
• a. Infection
• b. Oral hygiene
• c. Mucositis
• 3. Extraneous
• a. Gingivitis
• b. Plaque
• c. Use of antibiotics
• d. Body temperature
• e. Leukopenia

6

• E. How are the variables defined?
• 1. Rinsing solution
• a. Experimental Chlorhexidine-based solution
• b. Control Amine-stannous flouride
• 2. Infection C-reactive protein level
• 3. Oral hygiene Aerobic and anaerobic bacterial
  count
• 4. Mucositis Mucositis assessment tool
• 5. Gingivitis Loe and Sillness Index
• 6. Plaque Sillness and Loe Index
• 7. Use of antibiotics Comparison of remaining
  rinse solution and rinse diary record
• 8. Body temperature Degrees C
• 9. Leukopenia lt1000 Leucocytes/µL for at least
  3 days

7
2.1 Methods Design

• A. What design was used?
• 1. Experimental
• a. Random assignment to groups
• b. Manipulation of independent variable
• d. Control
• Exclusion criteria
• Sample stratified by gingivitis and plaque status
• B. Was it appropriate for the research question?
• 1. Yes
• a. Level 3 questions require an experimental
  design.

8
2.2 Methods Sample

• A. Appropriate for question?
• 1. Yes
• B. Target population?
• 1. Cancer patients who become leukopenic
  following chemotherapy
• C. Criteria for inclusion/exclusion?
• 1. Inclusion Cancer patients at least 18 years
  old who had normal oral mucosa, a normal body
  temperature, and a CRP level lt 50 mg/L and who
  provided written consent.
• 2. Exclusion Mucositis before chemo,
  intolerance to chlorhexidine or amide-stannus
  flouride, mental disorder, CRP 50 mg/L or temp gt
  38 C before chemo, removable dentures, diseases
  of upper respiratory tract

9

• D. How was the sample selected?
• 1. All patients receiving chemo with an expected
  leukopenia of less than 1000 leucocytes/µL
  treated at the Dept. of Haematology and Oncology
  from October 1999 to October 2001.
• a. Stopped after 47 patients due to an
  unexpected high number of severe mucositis cases.
• 2. Stratified according to plaque and gingivtis
  status (established tools but no R and V data
  provided).
• 3. Stratified block randomization to
  experimental or control group using a set of
  random sampling numbers
• a. Chlorhexidine group n24 (11 males, 4
  females)
• b. Control group n23 (10 males, 5 females)

10

• E. What was the sample size?
• 1. How was it justified?
• a. Power analysis
• 1.) 110 patients required to detect an increase
  in C-reactive protein (CRP).gt 50 mg/L with a
  power of 80 (alpha0.05).
• 2.) Note sample size calculations always based
  on statistical test that will be used to evaluate
  change in dependent variable for primary research
  question.

11
2.3 Methods Data collection

• A. How/when were the data collected?
• 1. Infection C-reactive protein level
• a. During leukopenia
• 2. Oral hygiene Aerobic and anaerobic bacterial
  count
• a. 3 time points - before leukopenia, during
  leukopenia, and after leukopenia.
• 1.) Swab 1 places in 5 ml tryptic soy broth
• 2.) Swab 2 placed in 5 ml thioglycolate
• 3.) Both swab vessels supplemented with a
  neutralizer to stop antimicrobial activity
• 4.) Serial dilutions prepared and plated on
  Columbia blood agar plates
• 5.) Swab 1 incubated aerobically
• 6.) Swab 2 incubated anaerobically
• 7.) Both swabs incubated at 37 C for 48 hours
• 3. Mucositis Mucositis assessment tool
• a. 3 time points - at the same time as oral
  hygiene swabs taken

12

• 4. Gingivitis Loe and Sillness Index
• a. Prior to randomization by one dentist
• 5. Plaque Sillness and Loe Index
• a. Prior to randomization by one dentist
• 6. Use of antibiotics Comparison of remaining
  rinse solution and rinse diary record
• a. During visits to patients for collection of
  mucositis and oral hygiene data (home visits?)
• 7. Body temperature Degrees C
• a. During leukopenia
• 8. Leukopenia lt1000/µL for 3 days
• a. Not Specified

13

• B. Were issues pertaining to reliability and
  validity adequately addressed?
• 1. plaque and gingivitis - no
• 2. mucositis - no
• 3. CRP - yes
• a. This endpoint was selected because in
  leukopenic adults, CRP levels above 50 mg/L have
  a ppv of 95 and a sensitivity of 100 for
  infection.

14

• C. Were the procedures adequately explained?
• 1. Yes
• a. Blood counts monitored
• b.. Patients instructed in rinsing
• c. Neither the patient nor the dentist knew the
  group assignment
• d. Nurses, physicians, and oncologists did not
  know the specific drug used in the oral rinses
• e. Both groups received Amphomoronal rinse to
  prevent candidiasis.
• f. Rinses in brown bottles
• g. Once patients became leukpenic, stopped
  brushing teeth and started rinsing.
• h. Rinsed 3 times per day (30 minutes after meals
  or 0800, 1300, 1800 during parenteral nutrition)
  from the beginning of chemotherapy to the end of
  leukopenia.
• i. Used 20 ml of solution
• j. Rinsed for 30 seconds
• k. Noted rinsing in booklet
• l. Patients were assessed for aerobic and
  anaerobic bacteria and mucositis before rinsing,
  on day 2 of leukopenia and after leukopenia
  resolved.
• m. Clinicians checked volume of rinse remaining
  and compared it to number of rinses reported.

15

• Was the data collection strategy appropriate for
  the design?
• 1. Yes
• a. Data collected strategy yielded data that
  could be used to answer the research questions.

16
2.4 Methods Data analysis

• A. Were the research questions answered?
• 1. Yes
• a. Analysis done twice
• 1.) All patients (experimental n24, control
  n23)
• 2.) Leukopenic patients only (experimental n15,
  control15)
• b. All participants appeared to follow the
  protocol correctly.

17

• c. Oral microbial count (Friedmans test)
• 1.) All patients
• a.) Patients who rinsed with chlorhexidine-based
  solution (n24) had a significant decrease in the
  concentration of aerobic bacteria (P0.042) and
  anaerobic bacteria (P0.008).
• b.) Patients who rinsed with amine-stannous
  flouride (n23) did not show a significant
  difference for either aerobic (P0.987) or the
  anaerobic bacteria (P0.651).

18

• 2.) For leukopenic patients
• a.) Patients who rinsed with chlorhexidine-based
  solution (n15) had a significant decrease in the
  concentration of aerobic bacteria (P0.054) and a
  non-significant difference in the concentration
  of anaerobic bacteria (P0.097).
• b.) Patients who rinsed with amine-stannous
  fluoride (n15) did not show a significant
  difference for either aerobic (P0.368) or the
  anaerobic bacteria (P0.607).

19

• D. Number of febrile patients
• 1.) All patients The number of febrile
  patients in both groups was similar.
• a.) Odds ratio was 1.07 (CI 0.27-4.22)
• 2.) Leukopenic patients
• a.) The number of febrile patients in both groups
  equal
• E. Infection
• 1.) All patients More patients with a CRPgt 50
  mg/L found in chlorhexidine group, but this was
  not significant.
• Odds ratio was 3.13 (CI 0.82-12.39)
• 2.) Leukopenic patients More patients with a
  CRPgt 50 mg/L found in chlorhexidine group, but
  this was not significant.
• Odds ratio was 3.50 (CI 0.55-24.61)

20

• F. Mucositis
• 1.) All patients More patients with severe
  mucositis found in chlorhexidine group. This was
  statistically significant.
• Odds ratio was 6.3 (CI 1.02 - 49.67)
• 2.) Leukopenic patients More patients with
  severe mucositis found in chlorhexidine group,
  but this was not significant.
• Odds ratio was 5.69 (CI 0.75 - 53.10)
• G. Conclusions
• 1.) Treatment with a chlorhexidine-based product
  did not provide the clinical benefit expected.
• On the contrary, risk of mucositis and clinical
  sequelae seemed greater (findings not
  significant).

21

• B. Was there sufficient power to support the
  statements made about the relationships between
  the variables?
• 1. No Needed 110 to determine influence of
  chlorhexidine rinse on systemic infection (CRP gt
  50 mg/L) but only collected 47 evaluable cases
• a. Stopped earlier due to adverse side effects
  of treatment in experimental group.

22

• C. Were the statistical tests properly used and
  conducted?
• 1. Yes
• a. SPSS 11.0 for Windows used to for graphic
  presentation and analysis.
• b. Friedmans test (non-parametric) used to
  examine differences in the concentrations of oral
  bacteria before, during, and after leukopenia.
• c. Influence of rinsing on temperature,
  mucositis and CRP analyzed by calculation odds
  ratios and corresponding confidence intervals
  using EPI-INFO.

23

• D. Internal validity
• 1. Were other possible explanations for the
  findings explored?
• a. Yes
• 1.) Concentration of chlorhexidine was higher
  (0.3) than other products (0.1-0.2).
• 2.) The chlorhexidine-based rinse also contained
  ethanol and hydrogen peroxide which might have
  inflamed tissue.
• 3.) Possible protective influence of amine
  stannous fluoride.
• E. External validity
• 1. Would the findings be applicable to other
  settings, times, people, places?
• a. Not addressed, but generalizability is
  limited because participants were not randomly
  selected and they all came from the same setting.

24
Implications for nursing practice, education,
administration without further research?

• Dont use chlorhexidine-based mouth rinses for
  prevention of systemic infections in leukopenic
  cancer patients.

25
Next logical research question

• Consider a study regarding the protective effect
  of amine-stannous fluoride.