Risk Factors for Unsuccessful Treatment and Complications With Negative Pressure Wound Therapy

  Abstract: The aim of this retrospective study was to identify risk factors related to unsuccessful treatment and complications with negative pressure wound therapy (NPWT). Methods. A consecutive series of patients treated with NPWT for wounds of various etiologies (n = 87) from 2005–2007 at a general hospital in a large city (Stockholm, Sweden) were assessed for risk for unsuccessful treatment and complications associated with NPWT. Results. Twenty-nine percent of the patients treated with NPWT had unsuccessful treatment results. The strongest risk factors associated with unsuccessful treatment were pressure ulcers (OR 4.6) or a positive culture for Staphylococci (OR 3.4). The complication rate was 21%, of which 14 patients had to terminate treatment. A positive culture for either Staphylococci or Pseudomonas was strongly associated (P = 0.001) with risk of complications during NPWT treatment. Patients with insufficient peripheral circulation in the extremities had a risk of both unsuccessful treatment and complications. Conclusion. The findings of the present study stress the importance of evaluating bacterial cultures and adequate antibiotic therapy when infection is suspected. The status of the patient’s peripheral macrocirculation in the lower extremities seems to have a significant impact on the risk of unsuccessful treatment or complications. Therefore, is it of great importance to evaluate peripheral circulation status before initializing NPWT.

Introduction

Modern wound care consists of many different dressing materials and techniques. Commercial negative pressure wound therapy (NPWT) has been used to treat wounds of varying etiology since 1995.¹ The therapy has proven to be safe, and the treatment results are equal to other wound care methods.^2,3 Nevertheless, in earlier studies, there are indications that NPWT results can vary among wound etiologies, and that patient related factors could have an impact on outcomes.⁴ Therefore, it is important to identify those patients who would likely benefit from the treatment, and more importantly, those patients who most likely would not.   Negative pressure wound therapy has become the gold standard for some diagnoses, such as open abdominal wounds^5-7 and dehisced sternal wounds, following cardiac surgery.^8,9 The effect of NPWT remains unclear for wounds of other etiologies, such as pressure ulcers, wounds of infectious origin, diabetic wounds, and traumatic wounds. Previous studies are questionable regarding their reliability and validity, due to problems related to sample size, methods of randomization, allocation concealment, and diverging results.^2,3,10 Another problem is that the previous studies show a discrepancy in outcome measures and lack evaluation of the clinical importance of the different measure methods used.   Adverse events and risk factors with NPWT have not been the primary focus of most studies. A systematic review by Vikatmaa et al² focusing on the effects and safety of NPWT, showed that infection, skin irritation, and pain during foam dressing changes, were adverse events associated with the treatment. The US Food and Drug Administration (FDA) received reports of 6 deaths and 77 injuries associated with NPWT during 2008–2009.¹¹ Bleeding was the most serious complication and caused all 6 deaths and 17 injuries. Vascular grafts, sternal and groin wounds, anticoagulant therapy, and foam dressing removal were factors that increased the risk of bleeding complications. The most common complications were infection and foam dressing retention.¹¹ It is not surprising that a high frequency of infection exists, as earlier studies have shown that fermentative gram-negative bacteria, such as Pseudomonas aeruginosa, decreased significantly, while Staphylococcus aureus bacterial load increased in the wounds during treatment with NPWT.^12-14 However, none of the previous studies included risk analyses of different wound types or patient-related factors.   The aim of the present study was to identify wound and patient-related factors associated with an increased risk of unsuccessful treatment and complications during treatment with NPWT.

Material and Methods

  This retrospective study included all consecutive patients treated with NPWT from 2005–2007 at a general hospital in a large city area (Södersjukhuset, Stockholm, Sweden). Ethical approval to conduct this study was obtained from the local Ethics Committee (2008/2023-31).   Inclusion criteria. Patients with any wound type from any of the hospital clinics who had been treated with NPWT were included in the study. The patients were identified from hospital data charts and by support from the manufacturer that had been tracking patients undergoing NPWT treatment.   Baseline. To define patient and wound-specific factors, all medical records were examined and entered into a predefined protocol by an experienced, specialized wound nurse. When uncertainties arose regarding the interpretation of the records, a general surgeon or an orthopedic surgeon was consulted to finalize the interpretation.   The protocol information included patient gender, indication for initializing NPWT, and age at the time of treatment. The patients were divided into 3 age groups: < 60 years, 60–74 years, and > 74 years.   The American Society of Anesthesiologists (ASA) physical status classification system is a system for assessing the fitness of patients before surgery.¹⁵ The ASA rating system has 6 status levels. The attending anesthesiologist performed and documented the ASA assessments in patients’ records. An independent anesthesiologist, not involved in the project, retrospectively evaluated and determined ASA classification in cases where data were absent in the patients’ records. The ASA scores were in the analysis divided into 2 groups: ASA scores 1 and 2 (healthy patients or with mild systemic disease), and ASA scores ≥ 3 (patients with moderate or severe systemic disease).   Wound etiologies were initially recorded, as stated in the medical records, and were divided into 5 groups for further analysis: 1) open postoperative wounds after either orthopedic or general surgery (eg, open abdominal wounds and postoperative infections); 2) wounds related to peripheral vascular disease, including diabetic wounds and arterial leg ulcers; 3) wounds due to primary infections, such as necrotizing fasciitis and erysipelas; 4) trauma-related wounds that could not be treated with primary closure; 5) pressure ulcers.   The medical preconditions noted were diabetes mellitus (Yes/No); current cardiovascular diseases (Yes/No); atherosclerosis in lower extremities, defined as an ankle-brachial pressure index < 0.8 or a toe pressure < 30 mmHg for nondiabetic, and 50 mmHg for diabetic patients (Yes/No); or ongoing dialysis treatment (Yes/No). Information on smoking habits (Yes/No) and current alcohol abuse (Yes/No) was also collected from the medical charts. All patients with missing data were classified as not having the disease or habit.   Treatment. The NPWT system used in the study was V.A.C.® Therapy (KCI, Inc, San Antonio, TX). Continuous subatmospheric pressure of 125 mmHg was applied. The dressings were changed 2 or 3 times a week, or more frequently, depending on the amount of fluid the wound produced (or when the canister was full). All wounds with necrotic tissue were surgically debrided as needed.   The indications for initializing NPWT were divided into 4 categories and were registered in the protocol: 1) control of infection and exudate; 2) preparation for other surgical intervention (eg, secondary suture or skin transplant); 3) optimizing wound healing when other options had failed; 4) treatment of open abdominal wounds.   The treatment of all patients was analyzed by viewing all patients’ data charts, as long as follow-up was possible at the hospital inpatient or outpatient departments, and treatment time was recorded. The follow-up time ranged from 24 months to 48 months. Records of all bacterial cultures taken during the wound healing process, and the treatment with NPWT, were registered.   Outcomes. Treatment results were abstracted verbatim from the charts and divided into 1) Successful treatment—wound much improved; healed wound; skin graft performed; secondary sutured; 2) Unsuccessful treatment—no improvement or increased wound size; infected wound bed; treatment interrupted due to complications. All forms of complications noted in the patients’ charts related to the treatment were registered (Yes/No) and categorized as: 1) complication related to the wound; 2) complication related to the patient’s mental and physical general health status; 3) complication related to technical issues with the NPWT equipment.

Statistical Analysis

  SPSS 17.0 for Windows was used for statistical analysis. Nominal variables were tested using the Chi-square test or Fisher’s exact test. The results were considered significant at P < 0.05. Logistic regression (LR) was used to study the associations between unsuccessful treatment and different risk factors. Two dependent variables were used: successful/unsuccessful treatment result and complication (Yes/No). The risk factors in the model were the variables gender, age, ASA score, medical preconditions, smoking and alcohol abuse, wound type, and bacterial culture (Tables 1, 3). The associations are presented as odds ratios (OR) with 95% confidence intervals (CI).   The model strategy was as follows. First, crude associations with each risk factor were studied in univariable models. Second, multivariable models were used to study the association, with adjustment for the other variables. Stepwise forward regression with entry testing based on the score statistics and removal testing based on the likelihood ratio statistics were used. The two-sided P values for entry and removal were set at 0.05 and 0.1, respectively. Thus, this procedure resulted in the final model showing the risk factors that were significantly associated with unsuccessful treatment or complications. No imputation was used for missing data. The Hosmer-Lemeshow goodness of fit test was used to examine if the final model adequately fit the data, with P > 0.05 indicating an acceptable fit.¹⁶

Results

  Patients. Ninety-two consecutive patients were assessed for eligibility; 87 were followed up regarding treatment results. Follow-up data were missing for 5 patients (5%).   A majority of treated patients were male (53, 61%). The median age was 68 years (range, 16–92). The etiology of the wounds was postoperative (n = 45), peripheral vascular disease (n = 11), infection (n = 9), trauma (n = 8), and pressure ulcer (n = 14). Sixty-three (72%) of the patients had some type of comorbidity.   All patients (N = 87) had swab culture results, and 18 (21%) also had biopsy results. No differences were found between the cultures and biopsy findings. Eighty-one (93%) of the wounds had a positive bacterial culture. The patients with negative cultures had been treated with antibiotics before the cultures were taken. All patients with positive bacterial cultures were treated with antibiotics after consulting the hospital’s infection specialist.   Treatment results. Sixty-two patients were classified as “successful treatment” (71%) and 25 as “unsuccessful treatment” (29%). The lowest rate of successful treatment was noted for the patients where the indication for treatment was optimized wound healing, when other options had failed (P = 0.03).   Patients with atherosclerosis of the lower extremities, patients with diabetes, wounds due to peripheral vascular disease, pressure ulcers, positive culture for S aureus, and positive culture for P aeruginosa, were tentatively associated with unsuccessful treatment in the univariable logistic regression analysis (Table 1). All other variables were nonsignificant. However, only pressure ulcers and a positive culture for S aureus were associated with unsuccessful treatment in the adjusted multivariable logistic regression model. Pressure ulcers had a 6.2 greater probability of unsuccessful treatment compared to wounds due to postoperative complications in the multivariable analysis.   Bacterial cultures. Two bacterial species were associated with significantly increased risk of treatment failure in the univariable model: S aureus (OR 3.4) and P aeruginosa (OR 4.4). In the multivariable model, a positive culture for S aureus remained independently associated with unsuccessful treatment (4.7 greater probability), but a positive culture for P aeruginosa was insignificant. Patients who had had a positive culture for both S aureus and P aeruginosa all experienced unsuccessful treatment with NPWT (P = 0.01). The Hosmer-Lemeshow goodness of fit statistics showed that the final model adequately fit the data (0.28).   Complications. Complications related to NPWT were noted in 18 (21%) of the patients (Table 2). The highest frequency of complications was noted for wounds associated with peripheral vascular disease 6/11 (55%). Interruption of treatment due to complications was noted in 14 patients, mainly due to wound complications, and patients’ perceived and reported deterioration in quality of life during treatment. Only 1 patient related their deterioration in quality of life to increased pain during treatment. All others reported the strain of repeated foam dressing changes in the operation theater, nonfunctioning devices, and being attached to a machine all day, as being overwhelming.   The univariable logistic regression analysis showed that patients with atherosclerosis in the lower extremities, with wounds due to peripheral vascular disease, with a positive culture for S aureus, or with a positive culture for P aeruginosa were preliminarily associated with complications during NPWT (Table 3). In the multivariate analysis, the statistical significance disappeared for all variables, except the presence of S aureus (OR 14.3) and P aeruginosa (OR 18.5), which remained a significant risk of the patient developing complications.   There was a confounding effect between patients with wounds due to peripheral vascular disease and a medical precondition of atherosclerosis in the lower extremities.   The Hosmer-Lemeshow goodness of fit statistics showed that the final model adequately fit the data (0.87).

Discussion

  This study shows that 29% of the patients treated with NPWT experienced an unsuccessful treatment result. The strongest risk factors associated with unsuccessful treatment were having a pressure ulcer or a positive culture for S aureus in the wound. This study also shows a complication rate as high as 21%, where 14 patients had to interrupt their treatment due to those complications. Risk factors for developing complications during treatment with NPWT were a positive culture for either S aureus or P aeruginosa in the wound. Patients with insufficient peripheral vascular circulation in the lower extremities had the highest frequency and a non-statistically significant increased risk for both unsuccessful treatment and complication risk. The results indicate the need for thorough evaluation of the patient’s peripheral circulation before initializing NPWT.   One of the strengths of this study is that the participants were selected from a consecutive series, thereby avoiding problems with selection bias. Another strength is that by reviewing the patients’ charts, interruptions in treatment due to deterioration of quality of life, could be observed. As quality of life has not been in focus in previous research, this finding is of significant relevance, as the treatment should not only be successful, but also endurable for the patient. As there is a lack of studies comparing the quality of life when using NPTW and conventional treatment, this is an important finding that warrants further investigation. Deterioration of patients’ quality of life is also described in a pilot study investigating this, and the conclusion was that the treatment may worsen quality of life for some patients.¹⁷ Another small study also showed that the group treated with NPWT had a significantly higher anxiety score than the control group.¹⁸ This indicates that there could be possible accompanying psychological effects with NPWT.   Research that relies on medical chart review is known to be associated with methodological and psychometrical difficulties. To reduce these risks, different strategies can be used, as shown by Gilbert et al.¹⁹ The present study adhered to the Gilbert et al strategy, with the exception of blinding the chart reviewer. To achieve greater accuracy, uncertainties in the interpretation have been discussed within the research group for consensus. Information regarding medical conditions can be regarded as reasonably reliable, since all physicians regularly noted these conditions, whereas registration of smoking habits and abuse of alcohol were often neglected in the medical records. The omission of information in the medical charts regarding comorbidities, interpreted as nonexistent in the study, could play an important role when analyzing the results and may result in an underestimation in the association between risk factors and treatment result and/or complications. Smoking was not significantly associated with unsuccessful treatment or complications in this study, despite the evidence for smoking being harmful to both acute and chronic wound healing.²⁰ The most plausible explanation for this is that the authors interpreted missing data as nonsmokers, and thus, underestimated the risk.   This study consisted of a small sample of patients without a control; thus, the conclusions should be carefully considered.   The wounds due to peripheral vascular disease, including diabetic foot ulcers and arterial leg ulcers, had the poorest outcomes among the various wound types. This contradicts the existing consensus that there is evidence of success using NPWT for treating diabetic foot ulcers.³ However, in the previous studies on diabetic foot ulcers, patients have had a sufficient peripheral macrocirculation or the circulation status has been unclear. It lacks evidence for the efficacy of NPWT of wounds when there is an existing peripheral macrocirculation deficit. Only 1 randomized controlled trial seeking to determine whether NPWT positively influences the healing of lower limb ulcers caused by peripheral vascular disease has been undertaken. The results of that study, although it ended in 2006, have not yet been published.²¹ This, of course, leads one to question why the results of the study have not yet been published.   The present study indicates that the status of the patient’s peripheral macrocirculation is of clear importance for the treatment result and should therefore be taken into consideration as a risk factor when deciding on initiating NPWT. The same findings were reported by Clare et al²² in a retrospective study that concluded that NPWT is not preferable for patients with insufficient peripheral circulation. Kairinos et al²³ further explained this in a study on tissue perfusion under the NPWT foam dressing. They reported that perfusion in the wound bed decreased rather than increased, as earlier studies suggested, contributing to even lesser circulation, and inhibition of wound healing.   The strong association between pressure ulcers and unsuccessful NPWT treatment indicates that a pressure ulcer may be a severe enough risk factor and should not be overlooked when considering NPWT. The evidence that NPWT is successful in treating pressure ulcers is ambiguous in the literature; for example, NPWT for a pressure ulcer in the pelvic region was not superior to conventional wound treatment in a small study by Wanner et al.²⁴ However, NPWT for these wounds still may have an advantage regarding exudate control, and may require fewer dressing changes, which may alleviate the strain on the patient and could enhance a patient’s quality of life during treatment.²⁵ When treating pressure ulcers, quality of life should be a concern, as well as a readiness for extremely long treatment times for both patients and clinicians.   The strong association between a positive wound culture for S aureus and unsuccessful treatment/complications is not surprising. Previous studies have shown that the S aureus bacterial load increases during NPWT.^12-14,26 An increased bacterial load in the wound could explain the poor treatment results, even if the patients were adequately treated with antibiotics.   The finding that a positive culture for P aeruginosa in the wound was strongly associated with unsuccessful treatment in the univariate model and for complications in the multivariate model in the present study is more difficult to comprehend. Logically, P aeruginosa should flourish in an occlusive and moist environment, as in the case of the wound treated with NPWT. However, previous research shows that NPWT decreases the bacterial load of P aeruginosa in the wound.^12-14 The poor outcome and risk for complications could not be explained by any baseline characteristics as there were no significant differences between the patients with positive cultures for P aeruginosa in the wound and those without (figures not shown). One theory is that P aeruginosa has an ability of forming biofilm in the wound and could be an inhibitor of wound healing during NPWT.²⁷   The indication for initiating NPWT also affected the treatment time, although with a statistically insignificant difference (figures not shown). Though the difference between treatment time and indication was not statistically significant, it is a clinically interesting observation that wounds without a clear treatment goal—other than just optimized wound healing—needed a 1-week (or longer) treatment time. Perhaps wounds with a clear treatment goal have more diligent follow-up than those that do not. This also highlights the importance of follow up, and that the 2-week follow up timeframe the manufacturer recommends, seems to be appropriate.²⁸   The finding that positive cultures for S aureus and P aeruginosa are associated with unsuccessful treatment results and complications reflects the importance of bacterial cultures and adequate antibiotic therapy when infection is suspected.   As the present study has shown, a patient’s peripheral macrocirculation status seems to have a strong impact on the risk of unsuccessful NPWT treatment or complications. Therefore, it is crucial to evaluate peripheral circulation in the lower extremities before initializing NPWT, and to examine the cut-off point for when deficits in circulation might impact the outcome. Larger, prospective studies on the impact of comorbidities on treatment results and complications are needed, especially concerning the inconclusive variables in this study (ie, smoking and diabetes).

Conclusion

  In the present study, 4 patients experienced degraded quality of life during NPWT. The impact on the patients must have been severe since the patients expressed this to the physician, who documented it in the medical charts and terminated the treatment in all of the cases. Deterioration of patients’ quality of life has been described and concluded that NPWT treatment may worsen quality of life for some patients. Patients treated with NPWT have significantly higher anxiety compared to controls, which indicates that there could be accompanying psychological effects with NPWT. However, since there are few studies on NPWT and therapy outcome for wounds of different etiologies, there is a need for further studies to evaluate the outcome of NPWT, and to identify which patient groups will benefit most from the treatment. An effective and gentle wound treatment can decrease suffering and increase quality of life for patients. Wound treatment that does not achieve treatment goals can lead to a risk for the patient, and should be avoided.

References

1. Argenta LC, Morykwas MJ. Vacuum-assisted closure: a new method for wound control and treatment: clinical experience. Ann Plast Surg. 1997;38(6):563–576. 2. Vikatmaa P, Juutilainen V, Kuukasjarvi P, Malmivaara A. Negative pressure wound therapy: a systematic review on effectiveness and safety. Eur J Vasc Endovasc Surg. 2008;36(4):438–448. 3. Xie X, McGregor M, Dendukuri N. The clinical effectiveness of negative pressure wound therapy: a systematic review. J Wound Care. 2010;19(11):490–495. 4. Wallin A, Boström L, Ulfvarson J, Ottosson, C. Negative pressure wound therapy: a descriptive study. Ostomy Wound Manage. 2011;57(2):22–29. 5. Acosta S, Bjarnason T, Petersson U, et al. Multicentre prospective study of fascial closure rate after open abdomen with vacuum and mesh-mediated fascial traction. Br J Surg. 2011;98(5):735–743. 6. Kaplan M. Negative pressure wound therapy in the management of abdominal compartment syndrome. Ostomy Wound Manage. 2005;51(2A Suppl):29S–35S. 7. Swan M, Banwell P. Topical negative pressure. Advanced management of the open abdomen. Oxford Wound Healing Society; 2003. 8. Fleck T, Gustafsson R, Harding K, et al. The management of deep sternal wound infections using vacuum assisted closure (V.A.C.) therapy. Int Wound J. 2006;3(4):273–280. 9. Fuchs U, Zittermann A, Stuettgen B, Groening A, Minami K, Koerfer R. Clinical outcome of patients with deep sternal wound infection managed by vacuum-assisted closure compared to conventional therapy with open packing: a retrospective analysis. Ann Thorac Surg. 2005;79(2):526–531. 10. Ubbink DT, Westerbos SJ, Nelson EA, Vermeulen H. A systematic review of topical negative pressure therapy for acute and chronic wounds. Br J Surg. 2008;95(6):685–692. 11. US Food and Drug Administration. FDA Preliminary Public Health Notification: Serious Complications Associated with Negative Pressure Wound Therapy Systems; 2009. 12. Khashram M, Huggan P, Ikram R, Chambers S, Roake JA, Lewis DR. Effect of TNP on the microbiology of venous leg ulcers: a pilot study. J Wound Care. 2009;18(4):164–167. 13. Lalliss SJ, Stinner DJ, Waterman SM, Branstetter JG, Masini BD, Wenke JC. Negative pressure wound therapy reduces pseudomonas wound contamination more than Staphylococcus aureus. J Orthop Trauma. 2010;24(9):598–602. 14. Moues CM, Vos MC, van den Bemd GJ, Stijnen T, Hovius SE. Bacterial load in relation to vacuum-assisted closure wound therapy: a prospective randomized trial. Wound Repair Regen. 2004;12(1):11–17. 15. Donegan DJ, Gay AN, Baldwin K, Morales EE, Esterhai JL Jr, Mehta S. Use of medical comorbidities to predict complications after hip fracture surgery in the elderly. J Bone Joint Surg Am. 2010;92(4):807–813. 16. Hosmer DLS. Applied Logistic Regression. 2nd ed. New York: John Wiley and Sons Ltd; 2001. 17. Mendonca DA, Drew PJ, Harding KG, Price PE. A pilot study on the effect of topical negative pressure on quality of life. J Wound Care. 2007;16(2):49–53. 18. Keskin M, Karabekmez FE, Yilmaz E, Tosun Z, Savaci N. Vacuum-assisted closure of wounds and anxiety. Scand J Plast Reconstr Surg Hand Surg. 2008;42(4):202–205. 19. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med. 1996;27(3):305–308. 20. Ahn C, Mulligan P, Salcido RS. Smoking-the bane of wound healing: biomedical interventions and social influences. Adv Skin Wound Care. 2008;21(5):227–236. 21. McCarthy M. Vacuum assisted closure for the management of ischaemic wounds in the lower limb. A randomised controlled trial and in vitro studies. London: University Hospitals of Leicester NHS Trust; 2006. 22. Clare MP, Fitzgibbons TC, McMullen ST, Stice RC, Hayes DF, Henkel L. Experience with the vacuum assisted closure negative pressure technique in the treatment of non-healing diabetic and dysvascular wounds. Foot Ankle Int. 2002;23(10):896–901. 23. Kairinos N, Voogd AM, Botha PH, et al. Negative pressure wound therapy II: negative pressure wound therapy and increased perfusion. Just an illusion? Plast Reconstr Surg. 2009;123(2):601–612. 24. Wanner MB, Schwarzl F, Strub B, Zaech GA, Pierer G. Vacuum-assisted wound closure for cheaper and more comfortable healing of pressure sores: a prospective study. Scand J Plast Reconstr Surg Hand Surg. 2003;37(1):28–33. 25. Spilsbury K, Nelson A, Cullum N, Iglesias C, Nixon J, Mason S. Pressure ulcers and their treatment and effects on quality of life: hospital inpatient perspectives. J Adv Nurs. 2007;57(5):494–504. 26. Madsen SM, Westh H, Danielsen L, Rosdahl VT. Bacterial colonization and healing of venous leg ulcers. APMIS. 1996;104(12):895–899. 27. James GA, Swogger E, Wolcott R, et al. Biofilms in chronic wounds. Wound Repair Regen. 2008;16(1):37–44. 28. KCI. V.A.C. Therapy. Kliniska riktlinjer. Referenskälla för användare. London: KCI Europe Holding BV; 2007. Ann-Mari Fagerdahl, RN; Lennart Boström, MD, PhD; and Carin Ottosson, MD, PhD are from the Department of Clinical Science and Education, Karolinska Institutet, Stockholm, Sweden. Johanna Ulfvarson, RN, PhD is from the Department of Neurobiology, Care Sciences, and Society, Karolinska Institutet, Stockholm, Sweden. Address correspondence to: Ann-Mari Fagerdahl, RN Department of Clinical Science and Education Södersjukhuset, Karolinska Institutet Södersjukhuset S - 118 83 Stockholm, Sweden ann-mari.fagerdahl@sodersjukhuset.se