Abstract
Objectives
To evaluate the level of concordance between the 2007 PRETEMED guidelines and the 2012 American College of Chest Physicians (ACCP) guidelines in medical patients at admission.
Methods
A cross-sectional, observational and descriptive study was designed and included all adult medical patients admitted from an emergency department. Firstly, patients classified as low-moderate risk and high risk according to PRETEMED were compared to those classified by ACCP as low and high risk. Secondly, the same analysis was performed but this time low and moderate-high risk patients according to PRETEMED were compared to ACCP low and high risk patients. The level of concordance was calculated using the kappa concordance index. The study was approved by the Ethics Committee for Clinical Research of the hospital.
Results
The analysis was performed with 207 patients; 53.1% were male and the median age was 75.3 years (minimum 18, maximum 100 years old). The most common diagnosis at admission was related to a respiratory disease (37.2%). The level of concordance was 0.59 (95% CI 0.48 to 0.70) when moderate risk patients were grouped with low-risk patients and 0.53 (95% CI 0.42 to 0.65) when moderate risk patients were grouped with high-risk patients.
Conclusions
The level of concordance between both guides is moderate. It would be helpful to confirm whether the level of agreement improves when the patient's condition stabilises after several days of hospitalisation.
Keywords: CLINICAL PHARMACY, ACCIDENT & EMERGENCY MEDICINE
Introduction
Venous thromboembolism (VTE) is a frequent cause of preventable illness and death in hospitalised patients.1 2
In 2008, the ENDORSE study, considered the largest study to address this issue, evaluated both the risk of VTE and the adequacy of thromboprophylaxis in 68 183 patients from 358 hospitals in 32 countries.3 4 This study revealed a large number of patients at high risk of VTE not receiving thromboprophylaxis according to the American College of Chest Physicians (ACCP) 2004 guidelines;5 it also showed that the use of preventive measures in surgical patients (58.5%) was more frequent than in medical patients (39.5%).
This lack of correlation between the degree of risk and the prophylaxis prescribed suggested that individual risk is not sufficiently valued, especially in medical patients who usually present, concomitantly, several risk factors for VTE.6 7
For this reason, numerous scientific societies have developed models to assess individual risk of VTE such as the 2007 PRETEMED guidelines (one of the first international guidelines to provide a table of well-defined risk calculations) and the Padua Prediction Score, included in the ACCP 2012 guidelines.8 9 Both risk-assessment models are the most accepted models in Spain (see online supplementary appendix 1). However, given the different risk category classification for VTE between the two guides, the level of concordance was assessed.
The aim of this study was to evaluate the level of concordance between the 2007 PRETEMED guidelines and the 2012 ACCP guidelines in medical patients at admission.
Methods
A cross-sectional, observational and descriptive study was designed and included all adult medical patients admitted from an emergency department (ED) during 7 days (between 20 November 2012 and 26 November 2012) at a 1400-bed tertiary care hospital.
Patients to whom anticoagulants were prescribed for therapeutic purposes, and patients admitted to the intensive care unit or maternity wards were excluded. Patients with contraindications for the prescription of pharmacological prophylaxis (hypersensitivity to the drug, background or suspected heparin-induced thrombocytopenia, active bleeding or increased risk of bleeding because of impaired haemostasis) were included in the study to assess their individual risk of VTE.
Variables collected
A database was designed and included the following information obtained from the ED medical record:
Demographical data (age, gender and bodyweight)
Diagnosis at admission
Analytical parameters at admission (international normalised ratio (INR), prothrombin time (s), activated partial thromboplastin time (aPTT) (s), fibrinogen (mg/dL), and number of platelets (×109/L)
Presence of risk factors for VTE (based on the PREMETED guide): age >60 years; air travel >6 h; active inflammatory bowel disease; acute myocardial infarction (AMI); antidepressants; antipsychotics; aromatase inhibitors; bed rest >4 days; central venous catheter; chemotherapy; class III chronic heart failure (CHF); class IV CHF; chronic obstructive pulmonary disease (COPD) with severe decompensation; neoplasia; diabetes mellitus; hormonal contraceptives; hyperhomocysteinaemia; HIV; hormone replacement therapy; lower limb injury without surgery; myeloma with chemotherapy; nephrotic syndrome; obesity (body mass index (BMI) >28 kg/m2); paralysis of lower limbs; pregnancy/postpartum; deep vein thrombosis; prior superficial vein thrombosis; severe infection; smoking >35 cigarettes/day; stroke with lower limb paralysis; tamoxifen/raloxifen; thrombophilia; vasculitis (Beçhet/Wegener).
Presence of risk factors for VTE (based on the Padua Prediction Score): active cancer; previous VTE; reduced mobility; already known thrombophilic condition; recent trauma and/or surgery (<1 month); being elderly (≥70 years); heart and/or respiratory failure; AMI or ischaemic stroke; obesity (BMI ≥30 mg/kg2); acute infections and/or rheumatological disorder; ongoing hormonal treatment.
Two experienced clinical pharmacists in VTE working in the ED were in charge of introducing the data into the database. The following sources of information were revised for each patient: electronic medical record (available on the hospital website), analytical results from the ModuLab programme, and in-hospital drug prescriptions, collected using the Prescriplant electronic prescribing software available on the hospital website.
VTE risk assessment
The individual risk of VTE was calculated based on the PRETEMED guide that classifies patients into low risk (if the score is equal to 1–3), moderate risk (if the score is equal to 4) and high risk (if the score is >4).
In addition, the individual risk was also assessed according to the criteria established by the ACCP 2012 guidelines. In this case, the guide classifies patients into two categories: low risk if the score is <4 and high risk if the score is ≥47 (Table 1).
Table 1.
Venous thromboembolism risk assessment according to PRETEMED and ACCP guidelines
| Adjusted risk | PRETEMED 2007 | ACCP 2012 |
|---|---|---|
| Low | 1–3 | <4 |
| Moderate | 4 | – |
| High | >4 | ≥4 |
ACCP, American College of Chest Physicians.
Level of concordance
The level of concordance between both guides was calculated. Firstly, patients classified as low-moderate risk and high risk according to PRETEMED were compared to those classified by ACCP as low and high risk. Secondly, the same analysis was performed but this time low and moderate-high risk patients according to PRETEMED were compared to ACCP low and high risk patients.
The study was approved by the Ethics Committee for Clinical Research of the hospital. No additional risk was observed and the informed consent was a non-compulsory requisite.
Statistical analysis
This was a descriptive study that utilised descriptive statistics only. Continuous, normally distributed variables were expressed as mean (±SD) and non-normally distributed variables were expressed as median with IQR. Proportions were expressed as percentages with IQR.
The concordance assessment between both guides was calculated using the kappa (κ) concordance index.
All statistical analyses mentioned were performed using the statistical analysis package SPSS V.15.0. Chicago, SPSS Inc.
Results
A total of 393 patients were examined and 186 of them were excluded as they did not fulfil inclusion criteria: nine patients (2.3%) were transferred to the intensive care unit, 83 (21.1%) to a maternity ward, six patients (1.5%) were terminally ill and only required supportive care, nine patients (2.5%) were diagnosed with acute VTE, 11 (2.8%) with acute coronary syndrome, six (1.5%) with ischaemic stroke, and 62 patients (15.8%) required chronic anticoagulation (six patients for pulmonary embolism, 10 patients for deep venous thrombosis, 45 for atrial fibrillation, and one had a prosthetic cardiac valve).
The final analysis was performed with 207 patients; 53.1% were male patients and the median age was 75.3 years (minimum 18, maximum 100 years old). The most common diagnosis at admission was related to a respiratory disease (37.2%). Bodyweight was not registered in the emergency medical record of any patient.
The characteristics of the study population are summarised in table 2.
Table 2.
Description of general characteristics
| N (%) | |
|---|---|
| Demographical characteristics | |
| Gender (male) | 110 (53.1) |
| Median age (years) (Q1–Q3) | 75.3 (53.5–82.9) |
| Analytical parameters | |
| Platelets at admission (×109/L)±SD | 211.6±100.2 |
| INR at admission±SD | 1±0.1 |
| aPTT (s)±SD | 29.4±5.5 |
| Prothrombin time (s)±SD | 12.4±7.7 |
| Fibrinogen (mg/dL)±SD | 612.7±192.0 |
| Diagnosis at admission | |
| Respiratory disease | 77 (37.2) |
| Others | 39 (18.8) |
| Genitourinary disease | 28 (13.5) |
| Gastroenterological disease | 28 (13.5) |
| Signs and symptoms not associated with a clear disease | 18 (8.7) |
| Cardiovascular disease | 17 (8.2) |
| Risk factors for VTE according to PRETEMED | |
| Age >60 years | 128 (61.8) |
| Air travel >6 h | 0 |
| Active inflammatory bowel disease | 1 (0.5) |
| Acute myocardial infarction | 0 |
| Antidepressants | 29 (14) |
| Antipsychotics | 18 (8.7) |
| Aromatase inhibitors | 0 |
| Bed rest >4 days | 34 (16.4) |
| Central venous catheter | 0 |
| Chemotherapy | 6 (2.9) |
| Class III CHF | 21 (10.1) |
| Class IV CHF | 5 (2.4) |
| COPD with severe decompensation | 22 (10.6) |
| Diabetes mellitus | 42 (20.3) |
| Hormonal contraceptives | 1 (0.5) |
| Hyperhomocysteinaemia | 0 |
| HIV | 13 (6.3) |
| Hormone replacement therapy | 0 |
| Lower limb injury without surgery | 1 (0.5) |
| Myeloma with chemotherapy | 0 |
| Neoplasia | 28 (13.5) |
| Nephrotic syndrome | 0 |
| Obesity (BMI >28 kg/m2) | 26 (12.6) |
| Paralysis of lower limbs | 0 |
| Pregnancy/postpartum | 0 |
| Previous DVT | 5 (2.4) |
| Prior SVT | 0 |
| Severe infection | 131 (63.3) |
| Smoking >35 cigarettes/day | 5 (2.4) |
| Stroke with lower limb paralysis | 0 |
| Tamoxifen/raloxifen | 0 |
| Thrombophilia | 0 |
| Vasculitis (Beçhet/Wegener) | 0 |
| Risk factors for VTE according to ACCP | |
| Active cancer | 28 (13.5) |
| Previous VTE | 5 (2.4) |
| Reduced mobility | 34 (16.4) |
| Already known thrombophilic condition | 0 |
| Recent (≤1 month) trauma and/or surgery | 25 (12.1) |
| Elderly age (≥70 years) | 124 (59.9) |
| Respiratory failure | 99 (47.8) |
| Heart failure | 26 (12.6) |
| Acute ischaemic stroke | 3 (1.4) |
| Acute myocardial infarction | 0 |
| Acute infection | 134 (64.7) |
| Rheumatological disorder | 33 (15.9) |
| Obesity (BMI ≥30 kg/m2) | 26 (12.6) |
| Ongoing hormonal treatment | 1 (0.5) |
ACCP, American College of Chest Physicians; aPTT, activated partial thromboplastin time; BMI, body mass index; CHF, chronic heart failure; COPD, chronic obstructive pulmonary disease; DVT, deep vein thrombosis; INR, international normalised ratio; SVT, superficial vein thrombosis; VTE, venous thromboembolism.
Risk factors for VTE
The most prevalent risk factor in our study population regardless of which guide was used was infection, followed by being elderly. The third risk factor was diabetes mellitus (20.3%) in the PRETEMED guide and respiratory failure in the ACCP guidelines (47.8%). The complete list of risk factors is shown in table 2.
According to the recommendations of the PRETEMED guideline, 44% of the patients was at low risk of developing VTE, 16.4% at moderate risk, and 39.6% at high risk. Based on the ACCP guideline, 57% of the study population was at low risk and 43% at high risk.
Level of concordance
On the one hand, when low-moderate and high risk patients based on PRETEMED guidelines were compared to ACCP low risk or high risk patients, the level of concordance between both guides was 0.59 (95% CI 0.48 to 0.70).
On the other hand, when low and moderate-high risk patients according to PRETEMED were compared to ACCP low risk and high risk patients, the level of concordance between both guides was 0.53 (95% CI 0.42 to 0.65) (table 3).
Table 3.
Level of concordance between the two risk-assessment models
| ACCP | PRETEMED | |
|---|---|---|
| Low-moderate risk N (%) |
High risk N (%) |
|
| Low risk | 101 (48.8) | 17 (8.2) |
| High risk | 24 (11.6) | 65 (31.4) |
| ACCP | PRETEMED | |
|
Low risk N (%) |
Moderate-high risk N (%) |
|
| Low risk | 80 (38.6) | 38 (18.4) |
| High risk | 11 (5.3) | 78 (37.7) |
ACCP, American College of Chest Physicians.
Discussion
The level of concordance between both guides is moderate. The number of factors included in each guide and the weight given to each factor differ one from the other.
There are a number of conflicting factors that could contribute to an increased risk of VTE according to the PRETEMED guide.
Firstly, COPD is given a score of 3 and it is a very common factor in our study population (10.6%). Some studies showed that patients with COPD have an increased risk of VTE, especially during acute exacerbations requiring hospitalisation.10 11 However, the ACCP scale does not contemplate this factor.
Secondly, the PRETEMED guide considers elderly patients who are above 60 years of age, but this threshold is higher in the ACCP scale (≥70 years).
Thirdly, a higher number of risk factors are included in the PRETEMED guide compared to the ACCP guideline (33 vs 14). However, the formula used by the PRETEMED guideline in order to calculate VTE risk can only be applied if the patient has at least one precipitating process or a process with adjusted weight ≥2.
However, there are also several conflicting factors that could contribute to an increased risk of VTE according to the 2012 ACCP guide.9
The final VTE risk score can be directly calculated and it does not depend on the existence of a specific process.
Both guidelines consider infection as a risk factor for VTE. Nevertheless, in the case of the ACCP, the infection does not need to be severe.
Common risk factors have a higher score in the ACCP than in the PRETEMED guide (eg, neoplasia and previous VTE have a risk score of 3, while in the PRETEMED scale, the score given to both of them is 2).
Both guidelines consider trauma as a VTE risk factor. However, in the PRETEMED scale, trauma should only be considered if it affects lower limbs; in our study population, most traumas were cranioencephalic (24 cases).
In 2012, Gallardo et al published a study that evaluated the level of concordance between the PRETEMED and the 2004 ACCP guidelines.12 In this case, the level of concordance was high (κ=0.68 and 0.81). One of the main reasons that could explain this discrepancy is related to the risk factor ‘bed rest >4 days’ which is given a high score in both guidelines. In the study of Gallardo et al, VTE risk assessment was performed at some point during the patient´s period of hospitalisation, whereas in our study it was calculated at admission. For this reason, the factor ‘bed rest’ was present in 55.5% of the patients in their study compared to a 16.4% in our research.
Furthermore, Gallardo et al12 pointed out that PRETEMED criteria overestimates VTE risk compared to the 2004 ACCP guideline due to the high number of risk factors considered (59.6% of patients were at high risk according to PRETEMED vs 50% based on ACCP). However, in our study, the ACCP scale classified more patients at high risk (43% vs 39.6%). Regardless of which guide was followed, a lower percentage of patients were classified as high risk because, as it has been mentioned above, the presence of ‘bed rest’ was not as common as in the Gallardo et al12 study (the risk score given to ‘bed rest’ is 2 in the PRETEMED guide and 3 in the 2012 ACCP scale).
Both sets are widely distributed throughout our country, so one of the main questions that arise after analysing the data is: who is doing well? One possible cause for the lack of agreement is the fact that the 2012 ACCP guidelines focus on hospitalised patients, whereas PRETEMED was developed in order to take into account the ambulatory setting. Future studies should investigate which scale better identifies patients at risk.
Examples from the literature also show frequent discrepancies between clinical guidelines, for example, regarding the treatment of type 2 diabetes or lipid-lowering management.13 14 For this reason, in our opinion, the most important issues to be considered when selecting which recommendations to include in an institutional protocol are to evaluate the available evidence thoroughly and to have this evidence discussed by a broad panel of experts.
Regarding study limitations, the most frequent bias is the difficulty of validating the information obtained from medical records, as it may either be wrong or may not have been recorded. Furthermore, single-site case studies limit the ability to generalise the results. It would be preferable to experience similar situations in other centres in order to confirm our finding or, alternatively, accept that we are dealing with an isolated incident in our institution.
In conclusion, the level of concordance between both guides is moderate. It would be helpful to confirm whether the level of agreement improves when the patient's condition stabilises after several days of hospitalisation.
What this paper adds.
What is already known on this subject
Venous thromboembolism (VTE) is the most common preventable cause of hospital death.
Adequate thromboprophylaxis has reduced the rate of hospital-acquired VTE.
Models have been developed to assess individual risk of VTE.
What this study adds
The level of concordance between two widely accepted VTE guides is moderate.
Footnotes
Contributors: AdL-P: acquisition of data, analysis and interpretation of data, drafting the article and final approval of the version to be published. RG-S: acquisition of data, analysis and interpretation of data and final approval of the version to be published. ED-G, AC-G: acquisition of data, substantial contributions to conception and design, analysis and interpretation of data, revising the article critically for important intellectual content and final approval of the version to be published. CP-I: acquisition of data, analysis and interpretation of data, revising the article critically for important intellectual content and final approval of the version to be published. BM-A, PA-D: substantial contributions to conception and design, revising the article critically for important intellectual content and final approval of the version to be published. AH-A: revising the article critically for important intellectual content and final approval of the version to be published. JAA-L, MS-S: substantial contributions to conception and design and final approval of the version to be published.
Competing interests: None declared.
Ethics approval: Hospital General Universitario Gregorio Marañón.
Provenance and peer review: Not commissioned; internally peer reviewed.
References
- 1.Raffini L, Trimarchi T, Beliveau J, et al. Thromboprophylaxis in a pediatric hospital: a patient-safety and quality-improvement initiative. Pediatrics 2011;127:e1326–32. 10.1542/peds.2010-3282 [DOI] [PubMed] [Google Scholar]
- 2.Khorana AA. Venous thromboembolism prevention in cancer outpatients. J Natl Compr Cancer Netw 2013;11:1431–8. [DOI] [PubMed] [Google Scholar]
- 3.Cohen AT, Tapson VF, Bergmann J-F, et al. Venous thromboembolism risk and prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-sectional study. Lancet 2008;371:387–94. 10.1016/S0140-6736(08)60202-0 [DOI] [PubMed] [Google Scholar]
- 4.de Lorenzo-Pinto A, García-Sánchez R, Pascual Izquierdo C, et al. Impact of the ENDORSE study results on thromboprophylaxis prescribing patterns in medical patients attending the emergency department. Int J Clin Pract 2015;69:81–6. 10.1111/ijcp.12461 [DOI] [PubMed] [Google Scholar]
- 5.Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126(3 Suppl):338S–400S. [DOI] [PubMed] [Google Scholar]
- 6.Abdel-Razeq H. Venous thromboembolism prophylaxis for hospitalized medical patients, current status and strategies to improve. Ann Thorac Med 2010;5:195–200. 10.4103/1817-1737.69104 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kerbauy MN, de Moraes FY, Kerbauy LN, et al. [Venous thromboprophylaxis in medical patients: an application review]. Rev Assoc Med Bras 2013;59:258–64. 10.1016/j.ramb.2012.11.008 [DOI] [PubMed] [Google Scholar]
- 8.Medrano Ortega FJ, Navarro Puerto A, Vidal Serrano S, et al. Guía PRETEMED 2007 sobre prevención de enfermedad tromboembólica venosa en patología médica. http://www.sademi.com/actividades/publicaciones/guia_medica_pretemed_2007.pdf (accessed 25 Oct 2014).
- 9.Kahn SR, Lim W, Dunn AS, et al. Prevention of VTE in nonsurgical patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141(2 Suppl):e195S–226S. 10.1378/chest.11-2296 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Rizkallah J, Man SFP, Sin DD. Prevalence of pulmonary embolism in acute exacerbations of COPD: a systematic review and metaanalysis. Chest 2009;135:786–93. 10.1378/chest.08-1516 [DOI] [PubMed] [Google Scholar]
- 11.Goldhaber SZ. Risk factors for venous thromboembolism. J Am Coll Cardiol 2010;56:1–7. 10.1016/j.jacc.2010.01.057 [DOI] [PubMed] [Google Scholar]
- 12.Gallardo Jiménez P, Guijarro Merino R, Vallejo Herrera V, et al. [Assessment of venous thromboembolism risk in hospitalized medical patients. Concordance between PRETEMED guide and the recommendations of the VIII conference of the American College of Chest Physicians]. Med Clin (Barc) 2012;139:467–72. 10.1016/j.medcli.2011.07.022 [DOI] [PubMed] [Google Scholar]
- 13.Martínez-Castelao A, Górriz JL, Sola E, et al. About the discrepancies between consensus documents, clinical practice guidelines, and legal regulations in the treatment of type 2 diabetes. Nefrologia 2012;32:419–26. [DOI] [PubMed] [Google Scholar]
- 14.Isma'eel HA, Harbieh B, Tamim H, et al. Discrepancies between two lipid-lowering guidelines for CVD prevention in seemingly healthy individuals–case study Lebanon. Cardiovasc Diagn Ther 2013;3:129–36. [DOI] [PMC free article] [PubMed] [Google Scholar]